1. in_ernet_dli_2015_16670_2015_16670_Roamn-Britain-And-The-English-Settlements

BIRLA CENTRAL LIBRARY

PILANI (Rajasthan)

Class No: 942.01

Book No: C69R

Accession No: 50129

This book has been

graciously presented by

Seth G. D. Birla

THE OXFORD HISTORY

OF ENGLAND

Edited by G. N. Clark

THE OXFORD HISTORY OF ENGLAND

Edited by G. N. CLARK

Provost of Oriel College, Oxford

I. ROMAN BRITAIN AND THE ENGLISH SETTLEMENTS.

By R. G. Collingwood, late Waynflete Professor of Metaphysical Philosophy in the Univer-

sity of Oxford, and J. N. L. Myres, Student of Christ Church, Oxford, and University

Lecturer in Modern History. Second edition, 1937.

(Published)

II. ANGLO-SAXON ENGLAND, c. 550–1087.

By Sir Frank Stenton, F.B.A., Professor of Modern History, University of Reading. Second

edition, 1947.

(Published)

III. FROM DOMESDAY BOOK TO MAGNA CARTA, 1087–1216.

By A. L. Poole, President of St. John's College, Oxford.

IV. THE THIRTEENTH CENTURY, 1216–1307.

By Sir Maurice Powicke, F.B.A.

V. THE FOURTEENTH CENTURY, 1307–99.

By G. Barraclough, Professor of Medieval History in the University of Liverpool.

VI. THE FIFTEENTH CENTURY, 1399–1485.

By E. F. Jacob, Fellow of All Souls College, Oxford.

VII. THE EARLIER TUDORS, 1485–1558.

By J. D. Mackie, C.B.E., Professor of Scottish History and Literature, University of Glasgow.

VIII. THE REIGN OF ELIZABETH, 1558–1603.

By J. B. Black, Burnett-Fletcher Professor of History, University of Aberdeen. 1936.

(Published)

IX. THE EARLY STUARTS, 1603–60.

By Godfrey Davies, Member of the Research Staff of the Huntington Library. 1937.

(Published)

X. THE LATER STUARTS, 1660–1714.

By G. N. Clark. 1934.

(Published)

XI. THE WHIG SUPREMACY, 1714–60.

By Basil Williams, F.B.A., Professor of History Emeritus, University of Edinburgh. 1939.

(Published)

XII. THE REIGN OF GEORGE III, 1760–1815.

By Richard Pares, C.B.E., Professor of History, University of Edinburgh.

XIII. THE AGE OF REFORM, 1815–70.

By E. L. Woodward, Professor of Modern History, University of Oxford. 1938. (Published)

XIV. ENGLAND, 1870–1914.

By R. C. K. Ensor, Fellow of Corpus Christi College, Oxford. 1936.

(Published)

(The titles of some volumes are provisional)

ROMAN BRITAIN

AND THE

ENGLISH SETTLEMENTS

By

R. G. COLLINGWOOD, F.B.A.

Sometime Fellow of Pembroke College and

Waynflete Professor of Metaphysical Philosophy

and

J. N. L. MYRES

Student of Christ Church and

University Lecturer in Modern History

SECOND EDITION

OXFORD

AT THE CLARENDON PRESS

Oxford University Press, Amen House, London E.C. 4

GLASGOW NEW YORK TORONTO MELBOURNE WELLINGTON

BOMBAY CALCUTTA MADRAS CAPE TOWN

Geoffrey Cumberlege, Publisher to the University

FIRST EDITION 1936

SECOND EDITION 1937

REPRINTED 1941, 1945, 1949

PRINTED IN GREAT BRITAIN

PREFACE TO BOOKS I-IV

THIS volume is not a work of collaboration. It consists of two independent studies of two distinct, though interlocking, subjects. When the general scheme of the Oxford History of England was being discussed, the persons who undertook to write on the Roman and Anglo-Saxon periods agreed in thinking that the abstruse and delicate problems of the English settlement ought to be handled by some third writer, specially qualified to act as an intermediary between them, and able to give such a generous allowance of time and labour to the subject as should result in advancing our knowledge of it by an appreciable amount. We thought ourselves fortunate in being thus associated with Mr. Myres; and the fruit of his researches appears as Book V of this volume. It is printed here, not in the volume which deals with the Anglo-Saxon period, because the scale of treatment makes it easier to find room for it here.

Thus each part of the present volume is a work for which its own author is individually responsible. We have, in fact, deliberately refrained from discussing the connexions between our two subjects until our manuscripts were almost complete, believing that perfect independence of treatment was, in this particular case, the best way of arriving at the truth. Each of us was already familiar enough with the other's mind, methods, and ideas to be sure that the results would be reasonably harmonious. If the reader can detect contradictions between the two works, he may be certain that they do not arise from incompatibilities of temperament or divergences of prejudice on the part of the writers, but from those differences of perspective which any period of history presents when approached from two different sides.

Histories of England traditionally begin with some account of prehistoric ages. In this work that subject has been omitted. The invasion of Julius Caesar has been taken as the starting-point, and the book begins with an attempt to describe the state of the country as it was when that invasion took place. The reason for this does not lie in any theoretical distinction between history and prehistory, for that distinction does not here arise; from the beginning of the Neolithic period at least, British chronology is determined by strictly historical methods

vi

PREFACE TO BOOKS I-IV

through keying its archaeology into the chronological frame-

work given by documents from the Mediterranean region; it

lies in the fact that our knowledge of what is conventionally

called British prehistory is now so large, and is growing and

changing so rapidly, that it can no longer be dealt with in a

mere introduction. It demands a volume to itself, and the time

for writing that volume has not yet come.

In writing this study of Roman Britain, my aim has been to

make a contribution to the history of Britain, regarded as a

region with a personality of its own, and a part of its own to

play in the history of Europe. I have tried to reconstruct the

state of the country and of its people, and to determine how

far these altered, and how far they remained the same, in the

four centuries and a half with which I have dealt. Military and

political history are discussed only so far as seemed necessary

for this end. Conceived in this way, the Roman period is no

longer an isolated episode in British history. However isolated

it may seem when we regard it simply as an age when an alien

military and political system was imposed for a time upon a

distant province, it presents a very different aspect when we

reflect that the people upon whom this system was imposed

were people with an historical tradition of their own, who learnt

from Rome not the lessons she was able to teach but the lessc ns

their previous training enabled them to learn. I have tried to

write the history of Roman Britain as if it were the chapter in

a biography dealing with a man's school-days, in which he is

still the same boy that he was before he left home; whereas the

historian of the Roman empire is writing the history of the school.

A history of the Romano-British people has to be written with

attention to the somewhat meagre literary sources, eked out by

analogy from those of other provinces; but its material comes

mainly from archaeology. This country has been fortunate

in having a long and widespread tradition of archaeological

interest, and in the last half-century it has enriched and purified

this tradition by what is nothing short of a revolution in the

scientific method of archaeological research. It is the great

service of Haverfield to Romano-British studies that he set

himself, and to a great extent accomplished, the enormous

task of collecting and sifting the materials thus accumulated;

but death carried him off before he could digest the results

into a book. His various sketches, studies, and lectures are the

PREFACE TO BOOKS I-IV

vii

vindemia prima from which any book such as this must start.

My debt to other writers is, I hope, sufficiently acknowledged

in the bibliography. But this book could never have been

written without the co-operation of innumerable scholars who,

for the last twenty-five years, have helped me to form my ideas

by explaining to me their own. Every page, as I wrote it, has

recalled the generous friendship of these colleagues and the

happy relations with them which I have enjoyed.

Four special debts must be mentioned. My colleague

Professor J. R. R. Tolkien has helped me untiringly with prob-

lems of Celtic philology. Several years ago I found in Mr.

C. E. Stevens, now Fellow of Magdalen College, one who

shared my sense of the profound importance of agrarian prob-

lems for the social and economic history of the Roman empire,

and my conviction that hitherto the study of them had hardly

begun. His patient researches, which he has discussed with me

at every stage, have contributed more than I can hope to

acknowledge to my account of agricultural life in Britain.

Professor Tenney Frank honoured me by asking me to write

a section on Britain for his great Economic Survey of Rome, and

this detailed collection and discussion of economic material,

which will shortly be in print, he permitted me to use as the

groundwork of a great deal that appears in the present volume.

And Dr. R. E. Mortimer Wheeler has allowed me to read and

use in manuscript certain parts of his report on the late exca-

vations at Verulam, shortly to be published by the Society of

Antiquaries of London.

Lastly, I should like to express my gratitude to the University

of Oxford, which, by creating for me the post of University

Lecturer in Philosophy and Roman History, gave me for many

years together encouragement and opportunity to pursue the

studies whose results, having now vacated that position, I offer

to it and to the public.

oxford, 14 January 1936

R. G. C.

The second edition embodies corrections kindly supplied by

Mr. Robert Aitken, Mr. E. Birley, Dr. R. W. Chapman, Miss

Mary Kitson Clark, Sir Cyril Fox, Mr. E. W. Gilbert, Sir G.

Macdonald, Mr. I. C. Peate, Mr. C. E. Stevens, and Mr. R.

P. Wright.

16 February 1937

R. G. C.

PREFACE TO BOOK V

Between Roman Britain and Christian England there is a great gulf fixed, a void of confusion whose obscurity remains a standing challenge to historical inquiry. And not the least of the dangers which confront those who attempt to bridge the gulf lies in the difficulty of determining the limits of firm ground upon its borders. Chronologically, indeed, it may seem possible in certain directions to narrow the chasm almost to vanishing-point. In Book IV of this volume Professor Collingwood to our great advantage has pursued the flickering torch of Roman Britain into the gathering darkness of the sixth century, far beyond the conventional limit set for earlier writers by the edict of Honorius: and when his pursuit ends, he has in fact all but reached in time the solid land on which in the next volume Professor Stenton will lay the firm foundations of Christian England. It might thus seem that there is hardly room for a special study of the period between.

Nor have I attempted in Book V to study a period of years as such, but rather a cultural phase in our history which, however much we may learn to draw together the last days of Roman Britain and the first days of Christian England, will continue to be profoundly different from either and both to demand and to repay examination for its own sake. Book V is devoted to the establishment of the English Settlements in the lowland zone of our country, and, if limiting dates must be given for the process, I would say that it started soon before the middle of the fifth century and that by 593, the traditional year of Ceawlin's death, the political evolution of territorial states had advanced so far that a new age had begun. Within that century and a half the face of lowland Britain was transformed, and it is our ignorance of the course and character of that transformation which still makes the great gulf between Roman Britain and Christian England so difficult to cross.

This being so I have deliberately devoted my space to the discussion of those problems which are directly relevant to the settlement of the English in Britain. The surviving Celtic society of the west has been left on one side in the isolation to which the coming of the English, and its own preoccupation with the memories of Rome and with the specialized problems

PREFACE TO BOOK V

ix

of life in the highland regions, had forced it: it was not until after the end of our period that the conflict of religious enthusiasms over the soul of Anglo-Saxon heathendom brought Celtic society again into direct contact with English history. Nor has any attempt been made to discuss in themselves the character of Anglo-Saxon social life and organization, or the varieties of its artistic expression in the pagan period. The former subject will receive full treatment in the next volume, and, since the earliest sources on which our knowledge of its features depends belong almost wholly to Christian times, it would be not only redundant but possibly misleading to anticipate that discussion here. On the other hand, a review of the arts and crafts of pagan times would certainly be here in place: it has been reluctantly omitted mainly because it would be difficult, if not impossible, to make an exposition of the subject intelligible without the illustrations which the plan of the present series precludes.

The varieties of Anglo-Saxon zoomorphic ornament must be seen to be believed, and no amount of description can convey the fantastic beauty of the Kingston brooch. But while a separate discussion was also rendered impracticable by considerations of space, I have endeavoured to introduce the main conclusions that can be drawn at present from this cultural and artistic evidence for the character of pagan society wherever it seemed relevant to the general course of the narrative.

I have tried to indicate in the Bibliography and footnotes my debt to many scholars whose work in varied fields has contributed especially in the last few years to the growing illumination of the darkness. Two recent publications—Mr. R. H. Hodgkin’s History of the Anglo-Saxons, and the Ordnance Survey Map of Dark Age Britain—have rendered the paths of all students of this age immeasurably smoother than they were even two years ago, and my debt to both will be obvious. I owe it to them especially that the material on which my maps are based has been easy to verify, and that the means employed in its presentation have been easy to select: it may, however, be noticed that most of the archaeological entries on the maps had been independently collected by me before these works were published and that such differences as exist between my maps and theirs are the natural consequence of my own studies.

Where in my text I have ventured to disagree with Mr. Hodgkin or to criticize his interpretation of the evidence, it has been, I

x

PREFACE TO BOOK V

think, nearly always on topics on which he would be the first to

agree that difference of opinion is legitimate. Dr. S. W. Wool-

dridge’s important article in An Historical Geography of England

before 1800 (ed. H. C. Darby) was published after my chapters

were in proof: where our thought has run, as on many topics

it has, along parallel lines I can only congratulate myself on

our agreement, and regret that I was not able on some other

matters to profit earlier from his penetrating study. To Pro-

fessor Stenton I owe not only the inspiration which all readers

of his works derive from them but also much special help in

matters of philology and place-names, topics which have not

been my primary field of study. Mr. E. T. Leeds has kindly

assisted me in several archaeological difficulties, and has

generously allowed me to see the manuscript of his Rhind

lectures before their publication. Messrs. C. E. Stevens, C. F. C.

Hawkes, and P. C. Gordon-Walker have all been good enough

to read my proofs and have all made suggestions which have

saved me from errors of fact and ambiguities of expression.

To them and to my wife, who by typing the whole and mate-

rially assisting with the Index has taken on herself many of the

burdensome incidents of authorship, I would record my thanks,

for I appreciate to the full how much this work owes to their

encouragement and criticism.

J. N. L. M.

OXFORD

1 September 1936

CONTENTS

Book I. BRITAIN BEFORE THE ROMAN CONQUEST

I. THE STAGE OF HISTORY

The division of Britain into highland and lowland zones 1

Their differences in landscape, soil, climate 2

The permanent effect of the division on the history of Britain 3,

Its special effect on the Roman occupation 5

The highland zone as the military area 7

The lowland zone as the civil area 7

Human geography before the Roman conquest 8

The lowland zone 8

Chalk and oolite soils. Salisbury Plain as the centre of prehistoric Britain 8

The Neolithic Age and the origin of Agriculture 9

The Beaker invasion 9

The Early Iron Age 10

The contrast between the prehistoric and modern population-distributions 10

Sand and gravel soils 10

Uninhabited areas 11

Communications 13

The highland zone 13

Megalithic culture 13

The Beaker invasion 14

The Bronze Age 14

Communications between Britain and the Continent 14

II. BRITAIN IN THE TIME OF JULIUS CAESAR

Physical characteristics of the ancient Britons 16

Neolithic Age 16

Bronze Age 17

Roman period 17

The Romano-British type 17

Language 18

P-Celts and Q-Celts 18

Establishment of the Celtic language in Britain 19

Material civilization 19

Its Bronze-Age background 20

Late Bronze-Age civilization, its general character 21

Iron-Age cultures 21

(1) The earliest, of Hallstatt type 22

(2) South-western La Tène culture 24

(3) North-eastern La Tène culture 25

(4) The Belgae 26

Coinage and currency 29

xii

CONTENTS

General sketch of Britain in Caesar's time . . . 29

The south-east . . . 29

The Dorset-Yorkshire zone . . . 30

The western fringe . . . 30

The name Britannia . . . 30

III. CAESAR'S INVASION

Caesar's intentions and motives . . . 32

His reticence on these points . . . 32

His tacit admission of failure . . . 33

Origin of his British adventure . . . 33

His intention must have been conquest . . . 34

The reconnaissance of 55 B.C. . . . 34

Information sought from Gaulish merchants . . . 35

A fleet of transports extemporized . . . 35

Caesar sails for Britain . . . 36

The adventures of Commius . . . 36

Caesar's landing . . . 37

The gale and damage to his fleet . . . 38

Damage repaired: Caesar returns to Gaul . . . 39

Results of the reconnaissance . . . 39

Caesar's apologetic tone . . . 39

His failure to understand Channel navigation . . . 40

The gain in information small . . . 40

British tactics and the chariot . . . 40

British agriculture and the food-supply for an army . . . 41

The invasion of 54 . . . 42

New transports . . . 42

Concentration at Boulogne . . . 42

Mutiny in the camp: the fate of Dumnorix . . . 43

The Channel crossed . . . 43

Caesar's march into Kent . . . 44

The gale, and his return to the coast . . . 45

The fleet repaired and beached . . . 45

The problem of Caesar's anchorage . . . 45

Cassivellaunus brought into the war . . . 46

His resistance to Caesar's renewed advance . . . 47

Caesar's victory in a general engagement . . . 48

Operations north of the Thames . . . 49

Attack on his naval camp . . . 50

Submission of Cassivellaunus . . . 50

Caesar returns to Gaul . . . 51

Results of the invasion . . . 51

Its success in a military sense . . . 51

Its political success . . . 52

Financial considerations . . . 52

Did Caesar intend to return and complete the conquest? . . . 52

Why he did not do so . . . 53

CONTENTS

IV. FROM CAESAR TO CLAUDIUS

The conquest from Rome's point of view merely deferred . . 54

Developments in Britain . . 54

The Catuvellauni: their extension north-westward . . 54

Commius creates a new western Belgic realm . . 55

Its archaeology . . 55

Its development . . 56

Its civilization . . 56

Coinage of these two realms . . 57

Their fusion into one, in the time of Cunobelin us . . 58

Other British coinages: Iceni, Dobuni, Brigantes . . 58

Distribution of British coinage . . 59

Its style: original derivation from Macedonian models . . 61

Romanizing movement in first century A.D. . . 62

Art of this period . . 63

Pottery . . 63

Metal-work and the finer arts . . 65

Chronology of this work . . 65

The Belgic Britons: their taste in art rather Roman than Celtic 66

Celtic art in the north (Brigantia), west, and midlands . . 67

Economic life . . 68

Agriculture . . 68

Mining and metallurgy . . 68

Trade with the Continent . . 70

Roman intentions with relation to Britain . . 71

Augustus: the rumours of his intended British expeditions . . 71

His excuses for not conquering Britain: Strabo's explanations . . 72

British refusal to court his friendship . . 73

His real intentions . . 74

Tiberius . . 74

Gaius: his projects for conquering Germany and Britain . . 75

Book II. THE AGE OF CONQUEST

V. THE CLAUDIAN INVASION

Its motives . . 76

The military situation . . 77

The heritage of Gaius . . 77

The financial situation . . 77

The political situation . . 78

The expeditionary force . . 78

The plan of campaign based on Caesar . . 78

Mutiny at Boulogne . . 78

The invasion . . 79

Landing in Britain . . 79

Richborough . . 80

The situation in Britain . . 80

No organized resistance . . 81

Togodumnus and Caratacus . . 82

xiv

CONTENTS

Battle of the Medway . . . . 83

Halt on the Thames . . . . 84

Arrival of Claudius and occupation of Colchester . . . 85

The settlement of Britain . . . 85

The realm of Cunobelinus becomes a province . . . 85

Verulam becomes a municipium . . . 86

Treatment of Non-Belgic realms . . . 86

Cogidubnus and the Regni . . . 87

The Iceni . . . . 87

VI. CARATACUS AND BOUDICCA

Caratacus . . . . 88

His flight to the west . . . 88

Progress of the Roman conquest along radial lines . . . 89

London its strategic centre . . . 90

Advance in three columns . . . 90

Ostorius Scapula governor . . . 91

He finds the Brigantes and Silures troublesome . . . 91

Creates a frontier on the Fosse . . . 91

Disarms the tribes south-east of this . . . 92

Revolt of the Iceni . . . . 92

Ostorius attacks the Degeangli. . . . 93

The Brigantes revolt against Cartimandua . . . 93

The war between Ostorius and Caratacus . . . 94

Ascendancy of Caratacus in the west . . . 94

Ostorius plants a legion at or near Gloucester . . . 94

The decisive battle . . . . 95

Caratacus betrayed to Ostorius by Cartimandua . . . 96

Continued resistance of the Silures . . . 96

To Ostorius . . . . 96

To Didius Gallus . . . . 97

who again supports Cartimandua against her subjects . . 97

To Veranius Nepos . . . . 98

The coming of Suetonius Paulinus . . . 98

His campaign in north-western Wales . . . 98

Boudicca . . . . 99

Revolt of Boudicca . . . . 99

Death of Prasutagus and treatment of his heirs. . . . 99

Beginning of the revolt . . . . 99

Fall of Colchester . . . . 100

Rout of the Ninth legion . . . . 100

Suetonius reaches London and evacuates it . . . 101

Destruction of London and Verulam . . . 102

Defeat of Boudicca . . . . 102

Suppression of the revolt . . . . 102

Severity of the governor . . . . 103

The new procurator, Classicianus, opposes him . . . 103

Nero's special commission supports Classicianus . . . 103

The views of Tacitus . . . . 103

CONTENTS

VII. FROM BOUDICCA TO AGRICOLA

The later years of Nero . . . . . 105

Peaceful state of Britain . . . . 105

The revolt of Vindex . . . . 106

The civil wars . . . . 106

Insubordination of the legions . . . . 106

Fall of Cartimandua . . . . 107

Revolt of Civilis . . . . 108

Accession of Vespasian . . . . 108

Petillius Cerialis . . . . 109

Julius Frontinus . . . . 110

Julius Agricola . . . . 113

His character . . . . 113

His first campaigns . . . . 114

His Caledonian war . . . . 115

Perdomita Britannia . . . . 116

The sequel . . . . 118

VIII. THE MAKING OF THE FRONTIER

Trajan's reign . . . . 120

The Roman idea of a frontier . . . . 121

The fortified road . . . . 121

The continuous barrier . . . . 122

Domitian's frontiers . . . . 122

The Trajanic frontier in Britain . . . . 124

The Vallum . . . . 124

Its character . . . . 124

Views as to its purpose and date . . . . 125

It is not the Trajanic frontier . . . . 126

The Stanegate . . . . 126

Reorganization in Trajan's reign . . . . 127

The accession of Hadrian . . . . 128

His work in Britain . . . . 129

The prelude to it:

(a) a disaster in Britain . . . . 129

(b) frontier-work in Germany . . . . 130

Comparison between his British and German frontiers . . . . 130

Hadrian's Wall . . . . 131

The Vallum and its probable function . . . . 132

The building of the Wall . . . . 135

Its garrison . . . . 136

The Roman army in Britain at this time . . . . 137

Auxiliaries . . . . 138

Legions . . . . 139

Native fortification in Wales . . . . 139

IX. THE FRONTIER AFTER HADRIAN

The Antonine Wall . . . . 140

Peculiarities in its design . . . . 140

3720 1

b

xvi

CONTENTS

How it was garrisoned . . . . . 141

Economies in construction . . . . 142

Comparison with the contemporary 'outer limes' in Germany . 143

Reasons for these peculiarities . . . 143

Motives for the creation of the Antonine Wall . . 146

Consequent redistribution of military units . . 147

History of the Antonine Wall . . . . 148

Its building . . . . . . 148

The Brigantian revolt and Julius Verus . . . 149

Was the Antonine Wall involved ? . . . 150

Another British war and Calpurnius Agricola . . 150

The disaster under Commodus . . . . 151

Restoration by Ulpius Marcellus . . . . 153

Evacuation of the Antonine Wall . . . . 154

The civil wars of 193–197 . . . . . 154

Clodius Albinus . . . . . . 155

The Maeatian invasion . . . . . 157

Septimius Severus . . . . . . 157

Virius Lupus in Britain . . . . . 157

Severus repairs Hadrian's Wall . . . . 158

His wars in northern Britain . . . . 159

Book III. BRITAIN UNDER ROMAN RULE

X. THE MACHINERY OF GOVERNMENT

The Imperial officials . . . . . 161

Legatus Augusti pro praetore . . . . 161

Legati legionum . . . . . . 161

Legati iuridici . . . . . . 162

Procurators . . . . . . . 162

Their work . . . . . . . 163

Taxation . . . . . . . 163

Imperial property . . . . . . 163

Self-government . . . . . . 164

Colonies . . . . . . . 164

The municipium . . . . . . 165

Constitution, &c., of such towns . . . . 165

Smaller urban communities (vici) . . . . 166

Tribal res publicae . . . . . . 166

Their general character . . . . . 167

Their number and position . . . . . 167

The concilium provinciae . . . . . 169

The capital of Roman Britain . . . . 170

Colchester . . . . . . . 170

London . . . . . . . 170

The administrative position of York . . . 171

The Severan subdivision into two provinces . . 172

CONTENTS

XI. THE PEOPLE

Distribution of the population . . . . 174

Sources of information . . . . 174

The lowland zone . . . . 175

The highland zone . . . . 176

Prehistoric character of this distribution . . 177

Density of the population . . . . 179

Rise or fall in population during the Roman period . 180

In the lowland zone . . . . 180

In the highland zone . . . . 181

Immigration . . . . . 181

Analysis of epigraphic evidence . . . 182

Transplantations in the later Imperial age . . 183

The evidence of physical anthropology . . 183

XII. THE TOWNS

The town as the vehicle of civilization . . 186

Contrast between Graeco-Roman and northern views . 186

For Rome, to civilize Britain meant to urbanize it . 187

The precedent of Gaul . . . . 187

The rise of town-life in Roman Britain . . 189

Mushroom-growth under Claudius . . . 189

Destruction by Boudicca. . . . . 190

Reconstruction in more orderly fashion . . 190

Flavian town-planning . . . . 191

Characteristics of these towns . . . 192

Their Romanizing function . . . . 193

Resistance to this movement, and evidence of its decline . 194

New impulse with Hadrian's visit . . . 195

Romano-British town-life at its apogee . . 196

Number of towns . . . . . 196

Their size in acres . . . . . 197

Their population . . . . . 197

Their economic relation to the country-side . . 198

Their political and cultural function . . 199

Their buildings and streets . . . . 199

Workshops and dwelling-houses . . . 200

Baths, water-supply, amusements, temples, walls . . 201

The decay of the towns . . . . 201

Archaeological evidence for it . . . 201

The general decay of town-life in the Empire . . 203

Begins in the third century . . . . 203

Rostovtzeff's theory of its causes . . . 203

Checked by the work of Constantius Chlorus . . 206

Renewed in the fourth century . . . 206

XIII. THE COUNTRY-SIDE

Agriculture in Roman Britain . . . 208

Agriculture based on towns and forts . . 208

xviii

CONTENTS

Villas and villages

The relation between them

They imply two economic systems

Two types of field

The aratrum and caruca.

The distinction pre-Roman

The villa system

Its pre-Roman origin

Its growth under Roman rule

Its function as an instrument of Romanization

Its prosperity outlasts that of the towns

Distribution of villas

Their choice of sites

Their architecture

Origin of their plan

Size and degree of civilization

Economic functions

Land-tenure

Crops, live stock, produce

The villages

Their general character

Romanization

Rise or fall in their population

Legal status of their inhabitants

Condition of the peasantry

XIV. INDUSTRY AND COMMERCE

Industry

Sketch of the development of Romano-British industry

Effects of the conquest

Reaction of producers to the new conditions

Equilibrium reached in third century

Collapse at the end of the fourth.

Mining and minerals

Ownership and labour

Gold

Argentiferous lead

Mendips

Flintshire

Shropshire

Derbyshire

Yorkshire and Alston Moor

Tin

Copper

Coal

Iron

Iron-smelting

Ownership and organization

Metal-working industries

CONTENTS

Iron goods . . . . . 234

Bronze-working . . . . . 234

Goldsmiths and silversmiths . . . . 234

Glass . . . . . 235

Pottery . . . . . 235

History of the pottery industry . . . . 235

Triumph of large-scale production . . . . 237

Legionary and auxiliary potteries . . . . 238

Bricks and tiles . . . . . 238

Textiles . . . . . 239

Fulling-establishments in villas . . . . 239

Development of the woollen industry . . . . 239

Communications . . . . . 240

Roads: general character of the Roman system . . . . 240

Local traffic . . . . . 240

The imperial post . . . . . 241

Construction of roads . . . . . 241

Water-borne traffic . . . . . 241

Coinage . . . . . 241

Commerce . . . . . 242

Imports . . . . . 242

Exports . . . . . 243

Supposed export of wheat under Julian . . . . 243

Internal trade . . . . . 244

Town fora and country markets . . . . 244

Hawkers and pedlars . . . . . 245

Trade across the frontiers . . . . . 245

With Scotland . . . . . 245

With Ireland . . . . . 245

XV. ART

Celtic art in pre-Roman Britain . . . . 247

Its specialized and abstract character . . . . 248

Its high quality . . . . . 249

Inartistic character of Romano-British civilization . . . 249

The problems to which this contrast gives rise . . . . 250

Contrast between the La Tène tradition and the Graeco-Roman . 251

Impossibility of fusing these two traditions . . . . 254

Apparent exception: the Bath Gorgon . . . . 255

Its Celtic character . . . . . 255

Absence of British analogues to it . . . . 256

Metal-work . . . . . 256

Survival of Celtic style in the north . . . . 257

In Ireland and unconquered Scotland . . . . 257

Pottery . . . . . 257

Celtic tendencies in Castor and New Forest wares . . . . 257

Celtic revival under the late Empire . . . . 258

Hanging-bowls . . . . . 258

Conditions of this revival . . . . . 259

xx

CONTENTS

XVI. RELIGION

Contrast between artistic and religious conditions in Roman Britain 261

The worship of the official Roman gods . . . 261

Iuppiter. . . . . . 262

Mars . . . . . . 262

Mercury and others . . . . 263

Genii . . . . . . 263

Eastern religions . . . . 263

Celtic religion . . . . . 264

Local gods . . . . . 264

Sulis, Nodens . . . . . 264

Brigantia and the gods of the frontier district . 265

Celtic gods and Roman worshippers . . . 266

Romano-Celtic temples . . . . 267

Their lowland-zone distribution and semi-Roman character . 267

Celtic gods imported from the Continent. . . 268

Vitiris . . . . . . 268

Lack of unity in Romano-British paganism. . . 269

Christianity . . . . . 270

Legends and testimonies concerning its early growth . 270

The fourth century . . . . . 270

Bishops and councils . . . . 271

Archaeological evidences . . . . 272

Situation at the end of the century . . . 273

Book IV. THE END OF ROMAN BRITAIN

XVII. FROM SEVERUS TO HONORIUS

The early third century . . . . 274

The Gallic Empire . . . . . 274

The age of reorganization . . . . 275

Piracy and the Classis Britannica . . . 275

Maximian as Caesar in Gaul . . . . 275

Carausius . . . . . . 276

Allectus . . . . . . 277

Constantius Chlorus in Britain . . . . 277

Reconstruction of the Wall . . . . 277

Towns rebuilt. . . . . . 277

Saxon shore fortified . . . . . 278

Diocletian's new imperial system . . . . 279

The Tetrarchy . . . . . 280

The Provinces. . . . . . 280

The Army . . . . . . 280

The new system in Britain . . . . 281

The House of Constantine . . . . 281

Constantine the Great . . . . . 281

Constans . . . . . . 282

Magnentius . . . . . . 283

CONTENTS

xxi

Constantius II . . . . . . . 283

Julian . . . . . . . 283

The House of Valentinian . . . . 284

The great Pictish war . . . . 284

Count Theodosius in Britain . . . . 284

Reconstruction of the Wall . . . . 285

Signal-stations in Yorkshire . . . . 285

Valentia . . . . . . . 286

Gratian . . . . . . . 287

Magnus Maximus . . . . . . 287

The end of the Wall . . . . . 287

The House of Theodosius . . . . 288

Stilicho in Britain . . . . . 288

item per lineam valli . . . . . 289

Cunedda . . . . . . . 289

XVIII. THE END OF ROMAN RULE

Constantine III . . . . . . 291

The Gothic war . . . . . . 291

Usurpers in Britain . . . . . 291

The rescript of Honorius . . . . 292

The controversy concerning a Roman reoccupation of Britain after

410 . . . . . . . 292

Continental writers . . . . . 292

Gildas . . . . . . . 293

Lack of archaeological evidence . . . 294

The evidence of coins . . . . . 295

The Notitia Dignitatum . . . . 295

Bury's view . . . . . . 295

Its breakdown over the case of the Wall . 296

The argument from the count of Britain . 297

Extent and character of the reoccupation . 299

The evidence of coins . . . . . 299

Dating of the reoccupation . . . . 301

XIX. BRITAIN IN THE FIFTH CENTURY

Condition of Britain early in the fifth century . 302

Effects of barbarian raids . . . . 302

Towns not destroyed . . . . . 302

Villas decayed . . . . . . 302

The nature of this decay . . . . 303

Money no longer much used . . . . 303

Peasant revolts and their consequences . . 303

Analogy with Gaul . . . . . . 304

Evidence from St. Patrick's life . . . 305

Growth of Christianity . . . . . 305

British affairs from 400 to 450 . . . 306

Tribal governments ruling the country . . 306

xxii

CONTENTS

The first visit of St. Germanus.

Incompetence of these governments

Economic state of the country

Spiritual conditions: Christianity

Pelagius . . . . . 308

Fastidius . . . . . 309

Ninian . . . . . 310

Caelestinus and Patrick . . . . 311

Ecclesiastical contact between Britain and Rome . . 311

The date of Easter . . . . 311

Pelagianism. . . . . 311

Germanus . . . . . 312

Decay of Irish piracy . . . . 312

Condition of Britain about 450 . . . . 313

The 'departure of the Romans': its effect on Romanization and

prosperity . . . . . 313

Vortigern and his times . . . . 314

A new political order in Britain . . . . 314

Efficiency of this order . . . . 315

Its lower moral and cultural standards . . . 315

The backwash of Celticism . . . . 316

The disappearance of Roman civilization . . . 316

Two extreme theories . . . . 317

The waste chesters . . . . 317

Evidence of evacuation but not of massacre . . . 318

No such evacuation in the highland zone . . . 319

Wars between Britons and Saxons . . . . 319

The wars of Ambrosius . . . . 319

Wansdyke and other military works . . . . 319

Arthur . . . . . 320

The testimony of the Historia Brittonum . . . 320

Interpretation of the testimony . . . . 321

Book V. THE ENGLISH SETTLEMENTS, by J. N. L. MYRES

XX. THE SOURCES FOR THE PERIOD: ANGLES, SAXONS,

AND JUTES ON THE CONTINENT

Importance of the period . . . . 325

Causes of its obscurity . . . . 326

Deficiencies of the literary evidence . . . 327

Unsatisfactory character of the archaeological evidence . 330

Evidence of place-names . . . . 332

A geographical approach essential . . . . 333

Angles, Saxons, and Jutes on the Continent . . . 336

Views of Bede and Procopius . . . . 336

Tacitus and Ptolemy on the North German tribes . . 338

Question of the Chauci . . . . 339

CONTENTS

xxiii

Saxon expansion and congestion . . . . 340

Increasing weakness of Roman frontiers . . . 342

Difficulty of distinguishing between Angles and Saxons . 343

Problem of the Jutes . . . . 345

Angles, Saxons, and Jutes in Britain . . . 347

Great tribal confusion . . . . 348

Inadequacy of Bede's distinctions . . . 349

The archaeological evidence . . . . 349

Summary: importance of other than tribal distinctions . 350

XXI. THE COURSE OF THE CONQUEST IN KENT AND

THE SOUTH-EAST

When did the Anglo-Saxon settlement begin? . . 352

Reasons for dating the Adventus Saxonum about the middle of the

fifth century . . . . 353

Difficulty of the Gaulish chronicles . . . 353

Uncertainty of the archaeological evidence . . 355

The literary tradition of the settlement of Kent . . 357

Its intrinsic probability . . . . 359

The distribution of Kentish cemeteries . . . 360

Their relation to that of the early lathes . . 361

Composite character of Kentish culture . . . 361

Its bearing on the Jutish problem . . . 363

The Jutes of Hampshire and the Isle of Wight . . 364

Their probable derivation from Kent . . . 364

Date of their settlement . . . . 365

Settlement of the South Saxons . . . . 366

The literary tradition . . . . 367

The place-names of Sussex . . . . 368

The South Saxon cemeteries . . . . 368

Settlement of Surrey . . . . 369

Evidence of archaeology and heathen names . . 370

Relation of Surrey to Middlesex . . . 371

The position of London in the age of settlement . . 372

Problem of the East Saxons . . . . 372

Condition of Essex in the Roman and Saxon periods: a backward

area . . . . . 373

Importance of the place-names . . . . 375

Contrast with Kent . . . . 376

The common features in the settlement of the south-east . . 377

Similarities of culture, institutions, and place-names . . 377

The extent of treaty settlement in the literary sources . . 378

Chronology of the story in Gildas . . . 379

Its relevance to the Anglo-Saxon traditions . . 380

Ælle of Sussex, the first Bretwalda . . . 380

Historical reconstruction of the age of settlement in the south-

east . . . . . 381

xxiv

CONTENTS

XXII. THE FENLANDS, EAST ANGLIA, AND THE PROBLEM

OF WESSEX

Importance of the fenlands in the Anglo-Saxon settlement .

. 383

Condition of the fens in the fifth century .

. 384

Early settlement of the valleys that converge on them

. 386

Comparison with Roman times .

. 387

Extent of settlement based on the fenlands .

. 388

East Anglia

. 389

Archaeological evidence for Teutonic occupation in the fifth

century

. 389

Probable fusion of early independent communities in the sixth

century

. 390

The expansion of East Anglia and the Cambridgeshire Dykes

. 391

Importance of the Icknield Way as a route to the upper Thames

. 393

The problem of Wessex

. 393

Character of the upper Thames valley .

. 394

Its early occupation by 'West Saxons'

. 395

Did they come from the Cambridge region?

. 395

Archaeological parallels with Middle Anglia

. 396

Conflict with the West Saxon traditions

. 396

Character of these traditions examined

. 397

Archaeological evidence for the settlement of Hampshire and

Wiltshire

. 399

Suggestion of a late occupation supported by the place-names

. 401

Was the later Wessex settled mainly from the north or south?

. 402

Meaning of the Wansdyke

. 402

The southern origin of the house of Cerdic

. 403

Who were the Gewissae?

. 404

Archaeological and institutional evidence for the settlement of the

middle Thames valley

. 405

Signs of a British enclave between the upper Thames and Middle

Anglia

. 407

Anglo-Saxon penetration of the Midlands

. 408

The Hwicce

. 409

Mixed Anglo-Saxon occupation of the Avon valley

. 409

Early exploitation of forest land in the Midlands

. 410

XXIII. THE HUMBRENSES

The Humber estuary a focus of early settlement

. 411

Probable extension to Mercia and Bernicia

. 412

Frisian connexions of the Humbrenses of Deira

. 412

The Roman cultural background here much weaker than in the

south

. 413

The settlement of Lindsey: its geographical isolation

. 413

The position of Lincoln in the period of settlement

. 414

The earliest Mercia, south of the Trent

. 416

Its northward expansion

. 417

The Yorkshire Wolds as the centre of Deira

. 418

CONTENTS

Early settlements round York . . . 419

Expansion northwards to the Tees valley . . . 420

Bernicia settled late and from the sea . . . 420

But the settlers came from the Humbrenses . . . 421

Struggles of the Bernicians with the Britons recorded by Historia Brittonum . . . 422

Character of the Bernician settlement . . . 422

Summary of the course of the Anglo-Saxon invasions . . . 422

Relation of the historic kingdoms to the earliest areas of settlement 423

Ceawlin and the later Wessex . . . 424

XXIV. THE CHARACTER OF THE CONQUEST

The problem of continuity between Roman Britain and Saxon England . . . 425

Recent increases of knowledge have not brought final answers . 425

Illustrations from anthropology and place-name study . . . 426

The problem of urban continuity . . . 427

Necessity of further topographical study . . . 428

Difficulties in interpreting the topographical evidence . 429

Literary evidence on problem of urban continuity . . . 431

Can we believe Gildas? . . . 432

Non-urban atmosphere of British and Saxon literary tradition . 434

The Christian Church and the fate of the towns . . . 435

Saxon attitude to town life . . . 436

The situation in Kent . . . 436

Case of London . . . 437

The problem of the country-side . . . 440

Breakdown of the villa-system . . . 440

Romano-British and Saxon villages . . . 441

Differences of agricultural equipment and method . . . 442

The contrast more complete in Wessex than elsewhere . . . 442

Interpretation of the Kentish evidence . . . 443

The survival of the Romano-British people . . . 444

Weakness of the case for extermination . . . 445

Anthropological opinion . . . 446

Native persistence in later laws of Wessex and Kent . . . 447

Evidence of Burial Customs . . . 448

The revival of Celtic Art . . . 450

Summary survey of the state of England soon after the middle of the sixth century . . . 451

Conclusion . . . 455

APPENDIXES TO BOOK V . . . 457

BIBLIOGRAPHY FOR BOOKS I-IV . . . 462

BIBLIOGRAPHY FOR BOOK V . . . 478

INDEX . . . 491

LIST OF MAPS

I. Roman Britain . . . . To face page 1

II. Villas and Villages in Hampshire and Wiltshire . " " 209

III. Chief Areas of Roman Mining in Britain . . " " 228

IV. (a) Local Cults of Celtic Religion; (b) and (c) Dedications to certain Celtic Gods of the Frontier Region . . . " " 272

V. The Continental Background to the English Settlements . . . Page 334

VI. The South-east of Britain . . . To face page 353

VII. The Fenland Settlement Area . . . " " 385

VIII. The Growth of Wessex . . . . Page 392

IX. The Humbrenses . . . . To face page 413

X. The Growth of the English Settlements . . " " 456

ROMAN BRITAIN

English Miles

0 10 20 30 40 50 60

Kilometres

0 10 20 30 40 50 60 70

Large towns

Legionary fortresses

Small towns

Forts

Villars

Signal stations

Roads

Dense woodland

Open woodland

Land over 1000 ft

for modern equivalents of place-names see index

MAP I. ROMAN BRITAIN

BOOK I

BRITAIN BEFORE THE ROMAN CONQUEST

I

THE STAGE OF HISTORY

THE country of Britain is divided by nature into two parts,

each with a character of its own, a complement and

contrast to that of the other. The north and west are formed

by an ancient mountain-chain, worn down by ages of erosion

into a plateau of hard rock, and this again has been deeply

scored by water and ice, and shattered by Atlantic weather,

until its seaboard has become ragged and irregular and its

western parts cloven by deep valleys with mountain-masses

rising steeply between them. Only in its eastern parts has it

retained the character of a continuous plateau, with long

stretches of uninterrupted high ground between one valley and

the next.

Leaning up against the eastern edge of this highland zone,

and sloping thence away to the east and south-east, lies a plain

composed of newer and softer rocks. These have been deposited

during successive ages in superimposed and more or less

parallel strata; but in the course of time the western highlands

have tended to rise a little out of the sea, carrying with them

the western edge of the plain, and its eastern edge has corre-

spondingly sunk below it, so that, if no erosion had taken place,

the plain as a whole would now slope gently upwards from

sea-level on the south-east to a height of two thousand feet or

more on the edge of the uplands; but, since erosion always acts

most powerfully on the highest ground, and so tends to reduce

a slope to a level by carrying that away, its effect has been to

remove the uppermost strata and expose those below them

differentially, most in the west and north, least in the south and

east. A traveller going from the Straits of Dover towards the

Welsh mountains, therefore, finds himself alternately rising in

a gentle slope, along the surface of one geological stratum, and

then, where this breaks off, dropping sharply down an escarp-

ment to the surface of the one next below it.

It is this differential erosion that determines the main lines

2

THE STAGE OF HISTORY

of surface-geology, and therefore of historical geography, in the plain of secondary rocks that forms the lowland zone of Britain.

For several reasons the contrast between these two zones is a fact of primary importance for British history.

First and foremost, the two differ widely in the general character of their landscape. The softer rocks of the lowland zone yield gentle slopes and wide expanses of almost level ground; the highland zone, though in scale and wildness its mountains are nowhere comparable with those of younger mountain-ranges, is everywhere rugged and broken in the aspect of its surface, steep in its slopes, and hardly anywhere even approximately level for more than a very little space.

There is also a general difference in the character of the soil. Almost all the lowland formations, however much they differ among themselves, have this in common, that they weather into fertile soils, naturally clothing themselves with a rich growth of turf or forest trees, and capable of being used by man either for arable farming or as grass-land. In the highland zone the soils are everywhere by comparison poor, at best thin and stony, at worst so sour with peat as to be useless, and by nature chiefly given over to scrubby, stunted trees and open heath.

In climate also there is a contrast. The highland zone, with inconsiderable exceptions, has a rainfall everywhere above the average (40 inches in the year) for the entire country; the low-land zone everywhere one lower; and indeed almost all the lowland zone, except in its south-western part, has on average less than 30 inches in the year. This difference, which would in any case have a great effect on the relative habitability of the two zones, is further increased by the fact that the parts of the lowland zone which receive most rain are the chalk and oolite uplands, which most quickly absorb it; so that few parts of the lowland zone, except actual fen-land, are ever so water-logged as most of the highland zone is for most of the year. And this water-logged condition, besides being in itself an enemy to human life, lowers the temperature and decreases sunshine by evaporation and by the formation of clouds. Thus the lowland zone as a whole is warm and dry, though nowhere too dry for the free growth of vegetation; the highland zone as a whole is cold and wet. But the contrast in general or mean temperature is modified by the fact that the warm Atlantic winds which bring rain to the highland zone diminish the intensity of winter

HIGHLAND AND LOWLAND ZONES

3

cold there, so that frosts arc more intense in the lowland zone,

especially towards the east coast.

These differences in respect of relief, soil, and climate between

the highland and lowland zones have deeply affected the life

of their respective inhabitants. In the lowland zone com-

munications, whether by land-routes or by boat along the

slow-flowing rivers, have always been easy; agriculture and stock-

farming, ever since man first introduced them, have prospered

on lands eminently suited to them; and neither heat nor cold,

neither drought nor excessive rainfall, has ever been an obstacle

to human prosperity. In the highland zone communication by

land is everywhere difficult, and water-transport, except by

sea, impossible; the character of the soil makes agriculture and

stock-farming alike a precarious matter, with small profits and

great liability to loss; and the cold, wet climate is a hindrance

to every kind of activity and a handicap to every form of

civilized life.

Hence, in degree, there has always been a general difference

between the richer, more comfortable, and more prosperous

life of the lowland zone and the harder and poorer life of the

highland; and there have been differences of kind as well,

arising out of the same causes. The lowland zone is not only

more accessible to invasion from the Continent, it is also more

attractive to invaders; because life is easier there, changes in

the way of living are less hard to bring about, and consequently

the history of the lowland zone shows from time to time pro-

found changes of this kind, partly through the coming of new

inhabitants, partly through the infiltration of new ideas. The

highland zone is unattractive to invaders, hard to invade, and

hard to conquer in detail when invaded; its landscape and

climate impose peremptory laws on any one, no matter whence

he comes, who settles there; all these causes, therefore, combine

to make it a region tenacious of its old customs, conservative

in temperament, stubborn to resist any kind of change. New

peoples and new ideas, when they gain a foothold in it, do so

at the price of compromise with the old; and the civilization

of the highland zone, analysed at any given moment in its

history, shows a curious blend of older and newer elements,

the old surviving though modified by the new, the new quickly

adapting themselves towards conformity with the old.

In spite of this contrast between highland and lowland zones,

4

THE STAGE OF HISTORY

the relation between the two is by no means one of mere difference. The sea which rings them round and holds them

together combines them into a single whole with a character of

its own. From the point of view of an observer on the European

mainland, Britain as a whole is deeply permeated by the charac-

teristics especially belonging to the highland zone: conservatism

of temperament, tenacity of customs, resistance to new ideas

except in so far as they modify themselves under the influence

of the old; and thus figures as the embodiment of a spirit of

compromise, sometimes admired for its cautious and practical

common sense and its loyalty to long-tried traditions, sometimes

condemned for backwardness in the march of progress, blind

devotion to lost causes, deficiency in logic, proneness to half-

measures and hypocrisy. These characteristics of the British

spirit arise from the fact that, in its relation to the Continent,

Britain in general plays the same part which in particular

belongs, within Britain, to the highland zone: cut off from the

mainland by the sea, it is difficult enough of access to form a

melting-pot in which new arrivals, whether of population or of

ideas, are assimilated and absorbed, conquering only by being

conquered.

For this reason the rhythms of civilization in Britain have

always been slower, and their pulsation less violent, than on the

European mainland. The quick and brilliant flowerings of

such continental movements as Celtic art, Roman imperial

civilization, medieval French architecture, Renaissance art and

thought, or the revolutionary spirit of the eighteenth century

have seldom failed to find a reflexion in Britain; but they

have been reflected there somewhat tardily and dimly, losing

much of their own specific quality, and never representing that

quality in its highest and intensest form, but achieving by

their assimilation in the life of Britain a certain solidity and

endurance, so that in compensation for their lack of a rapid

blossoming they are exempt from a rapid decay, and each phase

tends, far more than on the Continent, to overlap with the

next and to give an impression of relative permanence.

This slowness of rhythm, which at first sight makes Britain

a mere laggard in the movement of European history, a refuse-

heap on the edge of the ocean into which are swept the outworn

relics of ethnic migrations and spiritual fermentations, there to

linger indefinitely in a confused and incert mass, has nevertheless

DIFFERENTIAL ROMANIZATION

5

a positive function of its own. What it lacks in brilliance and

novelty, it makes up in a certain maturity or ripeness; the very

habit of blending together things so diverse as the Celtic spirit

and the Roman, the medieval and Renaissance, the feudal and

the democratic, has constantly produced new forms of life

and thought which Britain has given back to the Continent as

her own peculiar contribution to the life and thought of Europe.

In this way Britain has given to Europe the Irish and Anglo-

Saxon scholarship and art of the dark ages, the empirical

philosophy of the school of Locke, the principles of parliamen-

tary government, and the fruits of the industrial revolution.

And in many cases, within Britain itself, these new products

have arisen where the spirit of conservatism is strongest, in the

highland zone: the home of Irish Christianity, of Locke,

Berkeley, and Hume, of Trevithick, Watt, and Stephenson.

The history of Britain in the Roman period is primarily the

history of its partial conquest and occupation by Rome. From

this point of view the historian’s interest is focused on the ques-

tion how far Britain became romanized, and in the light of

what has already been said we should expect to discover a

partial and tardy romanization of the lowland zone, less

intense than the corresponding romanization of the continental

provinces, and penetrating little if at all into the highland zone.

And this in effect is what we find; but not the whole of it. There

are two other facts, less interesting to the historian of the Roman

empire, but more interesting to the historian of Britain.

First, the adoption of Roman civilization in the lowland zone

could take a lasting and stable form only at the price of assimila-

tion to the ways of life already established in Britain; and this is

the key to the difference which we shall find to exist between

two widely divergent phases of Romano-British civilization, one

characteristic of the earlier part of our period, one of the later.

Secondly, the infiltration of Roman civilization into the

highland zone, slow though it necessarily was, and hard to

detect in the scanty evidences of its actual occurrence, might

well leave behind it results more permanent and more momen-

tous for later history than the quicker and more spectacular

romanization of the lowland zone. This is the key to the part

played in British history by the heritage of Roman life and

thought after Britain had ceased to be a part of the empire.

So much for the larger features of British geography and their

3720.I

R

6

THE STAGE OF HISTORY

influence on history during the Roman period; we must now

study the geographical setting of that history in somewhat closer

detail.

The lowland zone, extending as far west as the base of the

Devonshire-Cornwall peninsula and the beginning of the Welsh

hills, and as far north as the Pennines and the uplands of county

Durham and Northumberland, is a single, undivided whole.

Its landward limits, clearly marked by the outcrop of the

palaeozoic rocks, may be represented by a line beginning at

Seaton, on the Devonshire coast, and running north to the

Bristol Channel, with an outlier of highland-zone character in

the Mendips; recommencing at the head of the Bristol Channel

and skirting the Forest of Dean and the Malvern hills to the

Wrekin, and thence bending westward and northward by

Shrewsbury, Oswestry, and Wrexham to the estuary of the Dee;

beginning again near Lancaster and passing round the edge of

the Pennines by way of Manchester, Derby, and Sheffield until

it reaches the sea near the mouth of the Tees. The lowland area

thus defined stands out on the map of Roman Britain as the

region of Roman towns and villas, the region of peaceful,

civilized, romanized life.

The highland zone is discontinuous; it is broken into three

parts by two deep inlets, where the western sea has penetrated

through it to meet the lowland zone, the Bristol Channel and

the bight of the Irish Sea on the coasts of Cheshire and

Lancashire. These three parts are very different in size. The

northernmost, including the whole of Scotland, the Border

counties, and the Pennine chain, is by far the largest as well as

the farthest removed from Rome; Wales, including the Welsh

marchcs, is much smaller, and the Dumnonian peninsula the

smallest of all. These three regions have, from the Roman point

of view, this in common, that they are all highland regions,

impenetrable in the main to civilizing influences; but their

isolation from each other made it possible to treat them differ-

ently, and their differences of size and position made such

differences of treatment inevitable. The Dumnonian peninsula,

however uncivilized, was too small and too easy of access to be

a focus of armed opposition to Rome: it never offered serious

resistance to conquest, and never received a military garrison.

Wales was another matter; and the map shows it as studded

with a network of forts. But for the most part these had done

DETAILS OF THE TWO ZONES

7

their work within a hundred years from the landing of the

Roman armies, and were abandoned as no longer useful; con-

quest was here complete. The Pennine-Scottish region was

never conquered in its entirety at all; such a conquest was for

a time projected, but soon abandoned; and a compromise was

reached by throwing a frontier-line across it and maintaining

a system of forts in the conquered portion.

Thus the distinction between civil and military areas in

Roman Britain, long familiar to historians, corresponds with

the distinction between lowland and highland zones, but with

modifications due to the division of the highland zone into three

unequal parts.

The three legionary fortresses at York, Chester, and Caerleon-

on-Usk served both as bases or focal points for the military

region and as links connecting it with the civil. This is because

all three lie on the natural frontier-line separating the highland

zone from the lowland, and, more particularly, at the three

points on that line where the lowland zone penetrates most

deeply into the highland, the three gates of the highland zone.

York, the natural capital of its own vale, stands where the

northernmost tongue of lowland country is squeezed out be-

tween the Pennines and the Cleveland hills. Chester commands

the gap between the Pennines and Wales, pushed as near as

may be to Wales, its primary objective. Caerleon similarly

commands the gap between Wales and the Dumnonian penin-

sula, but is thrust into the very throat of the Welsh mountains

as if to emphasize the fact that Devonshire and Cornwall needed

no military supervision.

The lowland zone, continuous though it is in extent, is more

variegated in character than a first glance at the map would

suggest. North of the Thames it is crossed over its entire extent

by two ridges of hill-country: the oolite limestone ridge that

begins as the Cotswolds and continues successively as the

Northamptonshire uplands and Lincoln Edge, and the chalk

ridge that begins in the Berkshire Downs, continues across the

Thames as the Chilterns, and curves round to the south side of

the Wash, reappearing in a continuation of the same curve

as the Lincolnshire Wolds and, across the Humber, as the Wolds

of Yorkshire. South-east of the Chiltern ridge, the chalk

disappears beneath the clay of the London basin; but it rises

again south of the Thames, emerging from under the tertiary

8

THE STAGE OF HISTORY

beds to form another chalk ridge, the Hampshire Downs and the North Downs of Surrey and Kent. But for the effects of erosion, it would rise higher still, to form a dome-like plateau covering the Weald; but all this has gone; the North Downs are broken off to southward by an escarpment, faced across the sandy and clayey Wealden forest-land by the corresponding escarpment of the South Downs, whence the chalk dips downward again to the Sussex and Hampshire coast. These three ridges of chalk are not parallel; they converge westward, and meet in Salisbury Plain.

Salisbury Plain has thus a peculiar significance for the geology and topography of lowland Britain. It is not only in itself a vast plateau of chalk, it is also a ganglion from which three chalk ridges radiate: one through the Chilterns to the Wash, one through the North Downs to the Straits of Dover, and one through the South Downs to Beachy Head. Yet another range of chalk downs branches off from its south-western edge and extends to the Dorset coast, while tangentially to it, separated from it only by a narrow gap about Chippenham and Trowbridge, runs the oolite ridge of the Cotswolds.

The importance of this ridge-system radiating from Salisbury Plain consists not so much in its elevation, which is in many places inconsiderable, as in the character of its soil. The chalk and the oolites are the only formations in lowland Britain yielding large, continuous tracts of land whose soil is at once reasonably fertile and naturally drained. Excessive moisture is the chief enemy that human life encounters in Britain; it is easiest for man to live, therefore, in places where fallen rain is quickly absorbed by a porous subsoil; and subsoils of this kind are provided by sands, gravels, chalk, and limestone. The importance of sands and gravels must be considered later on; here it is enough to say that in the extent and continuity of their deposits they are much inferior to the other two. An agricultural population in a somewhat primitive stage of development, unequal as yet to the task of draining the loamy soils, and therefore taking the line of least resistance by choosing those with a natural drainage, while at the same time civilized enough to appreciate the value of large continuous tracts of habitable land, where an effective political and religious organization could be developed, would fasten upon these plateaux and ridges of chalk and oolite as its natural home, and, as its corporate life took

SALISBURY PLAIN

9

shape, would tend to regard Salisbury Plain as its home-land par excellence, the economic, religious, and political centre of Britain.

Something of this kind did actually occur during the earliest ages of civilization in Britain. Settled agricultural life was introduced in the Neolithic age, during the third millennium before Christ; towards the end of that millennium great stone monuments were being erected in this country on a scale implying considerable wealth, a high degree of political organization, and a very definite system of religious beliefs. These megalithic monuments, there is good reason to think, were a fashion introduced by people sailing up the Atlantic coast from the Iberian peninsula and making their landfall in the south-west of Britain; many of them settled in Cornwall or worked their way by degrees northwards either to Wales and Scotland, where they settled upon the western headlands and islands, and so to the extreme north of Scotland, or else to Ireland, which is densely covered with their remains; but many others landed on the coast of Dorset, where the south-western chalk ridge reaches the sea, and established a great megalithic culture on Salisbury Plain and the Cotswolds. It is here that the megalithic civilization of Britain developed in its earliest, purest, and intensest form, as the evidence of dolmens and long barrows testifies; and it is here that, at a later date, when the cult of the stone circle came into existence, the greatest and most impressive stone circles were erected, at Avebury and Stonehenge.

The next great accession to the population of Britain was the arrival of the ‘Beaker people’, so called from the characteristic drinking-cup found in their graves, who came from the south-east; their remains are scattered along the eastern coast very much as those of the megalith-builders are scattered along the western. But in spite of the difference of origin there was to some extent an identity in their goal; on Salisbury Plain the two civilizations met and mingled. The same thing happened to a less degree elsewhere, notably on the Yorkshire Wolds and the limestone uplands of Derbyshire; but it is on Salisbury Plain and the chalk downs of Dorset that the evidence of fusion is greatest in quantity and most striking in quality; for at Stonehenge itself the stones appear to have been dressed with bronze tools, and bronze was unknown in Britain until the coming of the Beaker people.

10

THE STAGE OF HISTORY

In the Early Iron Age, once more, the chalk and oolite ridges maintained their importance; and once more it is on Salisbury Plain and the dry uplands adjacent to it that man has left the most impressive evidences of wealth and power in the gigantic hill-forts of the Plain itself, the downs of Dorset and Hampshire, and the Cotswolds. Thus, for two thousand years and more before the coming of the Romans, Salisbury Plain had been a centre to which every new civilization gravitated when it reached Britain, and at which it attained its highest expression of material and spiritual culture.

This ancient importance of the chalk and oolite plateaux as the regions of densest population, greatest wealth, and highest civilization comes to our minds to-day as a surprise. We are accustomed to see those regions as among the most sparsely inhabited in the country; and we are aware that this has been so for many centuries; long before the towns began to eat up the population of the country-side, we know, the medieval villages and the Anglo-Saxon villages before them tended on the whole to avoid those regions, or at most to penetrate them by creeping up their river-valleys, and to choose by preference soils of a heavier and more loamy kind. Here and there, it is true, a modern town or village, like Stow-on-the-Wold, with its Early Iron Age ramparts, inherits what may well be an unbroken tradition from prehistoric times; but these exceptions only serve to emphasize the general truth of the principle that the centres of population and wealth have shifted their ground since those times, leaving the chalk and oolite relatively bare.

This is not the place to consider when and why this change came about; it is enough here to call attention to its reality, in order to insist upon the fact that when the Romans came to Britain they found a population living not like the village population of to-day, mostly on heavy soils which before they could be used for agriculture had to be both cleared of timber and also drained of superfluous moisture, but mostly on light and naturally drained soils whose timber had been cleared by a gradual process during the Bronze Age, not so much by systematic felling as by the slow encroachment of grazing animals.

In addition to the chalk and oolite, naturally drained soils are afforded by various sands and gravels. The relation of early man to this group of soils is a complex matter. Whereas it is easy to generalize about the oolite- and chalk-lands, and to

EARLY MAN AND THE ALLUVIAL SOILS

accept the principle that early man settled everywhere on these

except where, as in the Chilterns, they were overlaid by super-

ficial deposits of heavy clay, no generalization will hold good

of the sandy and gravelly soils. Considerable tracts of them exist

in the lowland zone; for example, the New Forest, the tertiaries

of the Bagshot district, the greensand which crops out under the

chalk in the Weald and north of the Berkshire and Chiltern

escarpments, the dry, wind-blown sand of the Breckland border-

ing the Fens in Suffolk and Norfolk, and the richer brick-earth

that lies along the banks and terraces of the larger rivers. Both

in the Bronze Age and in the Iron Age, down to the coming of

the Romans, some of these tracts were densely inhabited and

others practically devoid of population; and this difference

cannot be correlated with differences in the quality of the soil.

Of the lightest and poorest, some, like the Wealden greensand

and the Bagshot and New Forest regions, were left unsettled;

yet the Breckland, among the poorest of all, was densely

populated in the earlier part of the Bronze Age, though in the

Iron Age this population was greatly diminished. Of the river-

banks and terraces, again, some were thickly settled in the

Bronze and Iron Ages. The most remarkable case is that of the

Thames, where finds from the Bronze Age onwards occur in a

continuous belt on both sides of the river from its mouth to the

Oxford district, where they spread out fanwise over the alluvial

soils from Dorchester to Eynsham. Hardly less striking is the

case of the Orwell, where an almost equally dense belt of finds

runs up the valley and merges into the thickly populated

districts of the Breckland and the Cambridgeshire chalk. The

nearest approach to a generalization would be to say that by

the time the Iron Age was established the poorest sands and

gravels, even where they had once been thickly inhabited, were

either unoccupied or occupied only by a small population,

whereas the richer alluvial soils along the chief rivers of the

south-east were densely populated, and increasingly so as time

went on.

Dense though it was, however, this population may have

been less advanced in wealth and political organization than that

of the chalk and oolite plateaux; it has nothing left comparable

with Stonehenge and Avebury or with the great hill-forts of the

upland regions. In part the difference may be due to lack of

materials; the valley-people had no stone for making such things;

12

THE STAGE OF HISTORY

they may have erected circles of timber, and they did sometimes construct river-side forts like the Dyke Hills at Dorchester. Yet the evidence, so far as it exists, does suggest a lower degree of civilization in the valleys throughout the prehistoric periods.

As opposed to these regions of primitive settlement, there were others which early man hardly touched at all. Some of these, composed of barren sand and gravel, have already been mentioned; but the largest in extent and most important to the historian are the heavy clay-lands which hold the water on their surface and were originally covered with damp oak-forest having a dense and almost impenetrable undergrowth. The most conspicuous of these are the Weald, the London basin, the boulder-clay plateau of East Anglia, the belt of Kimmeridge and Oxford clays that separate the Berkshire–Chiltern chalk ridge from the main ridge of the oolite, the Liassic clays of the midlands, and the Trias and glacial clays of Cheshire and Lancashire. To these may be added certain extensive clay-caps overlying the chalk, especially on the Chiltern plateau. All these were in prehistoric times either uninhabited, though doubtless visited by hunters, or else inhabited only here and there, where belts or islands of less intractable soil were to be found. There were also regions of marsh-land, habitable only where islands of drier ground emerged: the Fens, the marshes of Somerset, and the country round the head of the Humber.

These inhospitable regions were not only in themselves unpopulated; they also acted in varying degrees as barriers to the movement of civilization. Thus the Weald effectively cuts the South Downs off from communication with the North Downs and the Thames valley, with the result that whereas a civilization focused in Salisbury Plain can make itself felt along the southern chalk ridge as far as Beachy Head, one centred in the Thames basin leaves that ridge as a whole unaffected. Again, a people settled on the river-banks of Norfolk would tend under primitive conditions to live in a backwater of history, untouched by influences from the south and west. The Oxford and Kimmeridge belt seems to have been a forest-wall which Bronze Age influence coming from the south-east failed to penetrate, leaving the oolite ridge singularly poor in relics of that period. And, most important of all, the vast triangle of the sparsely inhabited midland forest, together with the cold and wet clay-lands of the Cheshire gap, made it difficult for any civilization

EARLY MAN IN THE HIGHLAND ZONE

rooted in south-eastern England to affect Wales or to find an

outlet on the Irish Sea.

Until the Roman road-builders set themselves deliberately

to break these barriers down, the main lines of communication

generally followed within the lowland zone were very different

from those of Roman and all post-Roman times. They fall into

two groups. First, there are the ridgeways that mark the lines

of dry upland: the Icknield Way, connecting Salisbury Plain

by way of the Berkshire Downs and the Chilterns with the

Wash; the Harroway running similarly along the North Downs;

and others along the South Downs and the Cotswolds. Secondly,

there are traffic-lines following the river-valleys, partly the

rivers themselves, carrying traffic in canoes, partly roads along

the river-banks. The Thames was an important waterway

throughout the Bronze and Iron Ages; so were the rivers that

flow into the Wash, and, farther north, the tributaries of the

Humber, especially the Trent and the Ouse.

For the highland zone a briefer survey will be enough. Here,

as a rule, the soils are impermeable; for even the carboniferous

limestone, which figures so largely in the Pennine region, is to

a great extent covered either with boulder-clay or with peat;

and consequently, though dry limestone soils are important

wherever they occur, for example in Derbyshire, in the York-

shire dales, and in parts of Westmorland, they are not the

decisive element in the distribution of early man. The most

powerful factor in that distribution was height above sea-level,

or rather, above the level of more or less permanent excessive

surface-water. Throughout the highland zone there are groups

and belts of megalithic remains, their distribution eloquently

attesting the sea-borne character of the civilization that pro-

duced them, on the lower shelves of land near the coast. In

Cornwall they occur especially near Land's End; in Wales

towards the headlands of Pembrokeshire and Carnarvonshire

and in Anglesea; there is a sprinkling of them on the Cumber-

land coast with a few penetrating up the Eden valley, and they

become commoner again in Galloway and round the Firth of

Clyde; then, after a considerable gap, they begin again in Skye

and the Outer Islands, and there are many near the northern

and eastern coasts of Caithness and in the Orkneys. Inland,

except for an important group on the limestone plateau of

Derbyshire, they are conspicuously rare. Taken as a whole,

14

THE STAGE OF HISTORY

they represent an agricultural civilization occupying somewhat low-lying, fertile, naturally drained soils, based on the sea, and derived from exploitation of the Atlantic traffic-route which connects the Iberian peninsula with the Biscayan coastlands, the peninsula of Brittany, and the western coast of Britain.

The Beaker people, coming as they did from the east and south-east, affected the highland zone very little except in the north. Here the eastern coast of Scotland from Inverness to Aberdeen, and the coastal plain of Northumberland, are densely studded with their remains, which conspicuously avoid the megalith region of Caithness, but except for that northern extremity occur freely all over the eastern coastland. But the Beaker civilization penetrated much more adventurously than the megalithic into the heart of the country, occupying the upland valleys of eastern Scotland and Northumberland, and thence working its way to the open and well-drained basin of the Eden valley. The mingling of megalithic and Beaker civilizations, which in the lowland zone we have found to be characteristic of Salisbury Plain, seldom appears, and never on a comparable scale, in the highland zone: it occurs once in Derbyshire, once in the Eden valley, and again in a region which, though geologically belonging to the lowland zone, is culturally ambiguous as between that and the highland, the Yorkshire Wolds.

Taking the Bronze Age as a whole, and setting aside northern Scotland as outside the main scope of this volume, we find a population which avoids the lowest lands and the heaviest soils as too wet and too densely timbered for occupation, and avoids the highest as too cold and exposed, but settles freely in the upland valleys and on shelves of land and foothills between the mountains and the valley-bottoms, especially where these have a permeable subsoil. Settlement is particularly dense in eastern Northumberland, with penetration up the Tweed and through the Tyne gap into the Eden basin; the dales of the central Pennines, especially Airedale, which carries a similar penetration over the watershed into the Ribble valley; the Derbyshire uplands; North Wales from Flintshire to Anglesea; western Cornwall; and the plateau of Dartmoor.

It remains to consider the natural lines for sea-borne traffic connecting Britain with the Continent, and to ask how these in their various ways affected its early history.

SEA-ROUTES TO BRITAIN

15

The first is the Atlantic route to Brittany and Spain. This impinges on Britain at its south-western extremity and there divides into two branches, one going north into the Irish Sea and so to the west coast of Scotland, the other east, up the English Channel. In prehistoric times the northern branch was the more important of the two; the eastern seems to have been of value only because it gave direct access to the Dorset coast and thus to the rich hinterland of Salisbury Plain. As long as Salisbury Plain was the heart of civilization in Britain all roads led to it; and thus, in the megalithic period, it acted as a magnet drawing sea-borne traffic from the Atlantic coastal route as far up-channel as Dorset. Apart from this deflection, the Atlantic route belonged at its British end to the highland zone; it was the pathway by which early civilization reached Cornwall, Wales, Ireland, and the west and north of Scotland.

The second route, or group of routes, crosses the Channel between Normandy and Hampshire, again with a strong tendency to be deflected towards the Dorset coast. These routes seem to have increased in importance during the Bronze Age and the earlier phases of the Iron Age; they were the main agents in developing the Late Bronze Age and La Tène I1 civilizations of the Wessex uplands.

The third group crosses the Straits of Dover and reaches Britain at the Kentish ports, the Thames, and the rivers of Essex. This traffic-line seems first to have become important in the Early Iron Age, when it introduced a civilization of Hallstatt type into south-castern England; from the first century before Christ it became, and remained throughout the Roman period, the most important of all.

Lastly, a fourth group crosses the North Sea from the Low Countries and Denmark. These routes were of great importance about the beginning of the second millennium before Christ as introducing the Beaker people, and again some twenty-five centuries later as the highways of Anglo-Saxon immigration; between these two dates they were of little consequence.

1 The pre-Roman Iron Age on the Continent may be conveniently regarded as occupying the last millennium before Christ, of which the first half is the Hallstatt period, the second the La Tène period, both called after sites at which their characteristic civilizations were first extensively studied, and both subdivided into minor periods. English antiquaries at one time gave to the La Tène civilization the name 'Late Celtic'; this name is now little used.

II

BRITAIN IN THE TIME OF JULIUS CAESAR

ANY attempt to give an account of Britain, as it was when Julius Caesar's invasion brought it for a moment into the

light of history, must begin with the question: to what race did the Britons of that age belong? or, in less ambiguous terms, since

the word race is apt to carry misleading associations, what kind of men were they in bodily form and habit?

This is a question which, in the present state of knowledge, cannot be answered briefly and decisively; but something can

be done towards clearing away misconceptions and indicating the direction in which an answer should be sought.

In the last century students tended to approach the question from a point of view partly archaeological and partly literary.

From skeletons found in the communal burial-chambers of long barrows, they reconstructed the picture of a short, long-headed, lightly built, Neolithic race, which they connected with

the similar peoples of the western Mediterranean. The beaker burials of the earliest Bronze Age gave them a strongly con-trasted physical type, tall, massive, and round-headed; and they

tended to think of this type as characteristic of the Bronze Age. Finally from the historical writers of the Roman age they

derived an idea of the Britons in Caesar's time as a race of tall, fair warriors of Celtic speech.

All these conceptions had in them a certain degree of truth, but all must to-day be qualified; and the result of these quali-fications is a picture of the ancient Britons which, if more

complex in some ways, is simpler in others.

In the Neolithic period, which is as far back as we need attempt to trace the British pedigree, the whole of western

Europe was inhabited by a race of long-headed men, varying somewhat in appearance, and especially in colouring, since

they were probably always fairer in the north and darker in the south, but in most respects substantially alike. Into this area of

long-headed population there was driven a wedge of round-headed immigrants from the east, known to anthropologists as

the Alpine race. Most of the peoples that have invaded Britain, from whatever direction they have come, have belonged to the

western European long-headed stock, and have therefore borne

THE ROMANO-BRITISH PHYSICAL TYPE

17

a general resemblance to the people already living there; and consequently, in spite of the diversities among these various

new-comers, the tendency in Britain has been towards the establishment and maintenance of a tolerably uniform long-

headed type. The chief exception to this rule is the Beaker people, who, whatever their precise origin, derived from an

Alpine strain in their ancestry a strongly brachycephalous type of skull; but the total influence of this strain upon the physique

of the Britons has not been great; and in estimating its importance we must bear in mind that, although the Beaker people

buried their dead unburnt, so that their physical characteristics are well known to the modern archaeologist, the Bronze Age

to which their coming was the prelude was in general an age of cremation: hence very little is known concerning the physical

type of the population in that period as a whole. Other exceptions are the Romans themselves, using that name in a narrowly

geographical sense for inhabitants of central Italy, and the Normans: these were both round-headed peoples. But apart

from these exceptions (of which the Roman may be almost ignored, for hardly any skulls attributable to that type have

been found in this country) practically all the skulls found in Britain, of whatever age, are of a long-headed type whose

closest affinities are with the ancient peoples of north-western Europe. This applies even to the skulls found in long barrows,

which are more akin to the Nordic long-headed type than to the Mediterranean.

It is not until the later Roman period, when cremation ceased to be practised and inhumation became general throughout the

country, that we can assemble a sufficient mass of evidence to give us a trustworthy general idea of the physique prevalent

in ancient Britain. When this happens, we find that the physical type is fairly constant. The head is moderately long, with a

flattish top, giving an upright, square, and somewhat low forehead, generally marked by a transverse groove above the eye-

brows: the back of the head projects strongly; the cranial capacity is about the same as that of an average modern

Englishman, the stature somewhat less; the figure is as a rule sturdy and muscular.

This Romano-British type, as anthropologists call it, is so widely prevalent in the later Roman period that it may safely

be regarded as having established itself before the Roman

18

BRITAIN IN THE TIME OF JULIUS CAESAR

period began. Its closest connexions on the Continent are with

Scandinavia, the Low Countries, and northern France; and it

has survived on a considerable scale in modern England. Sir

Arthur Keith, the chief living authority on this subject, raises

the question:1 ‘Do people . . . of the Romano-British type occur

in our modern population? I would answer that inquiry with

a confident Yes—in very considerable numbers, particularly

amongst the English middle classes.’

Beyond asserting the wide prevalence of this type, certainly

in the later Roman period and probably at its beginning, little

can be said here about the local and other variations which

undoubtedly existed. It is possible, for example, that the

modern prevalence of fair colouring in the east and dark in the

west is not altogether due to the Anglo-Saxon invasions, but is

connected with a much older division between north European

influences on the east of Britain and southern influences on the

west, coming by way of the Atlantic coast. It is also likely that,

in the south-east, the latest comers, whom the Roman invaders

would encounter as a military aristocracy, would be fairer than

the general bulk of the population, with whom they had not

yet been assimilated.2

In language, all the various British tribes with whom the

Romans came into contact, and probably all the inhabitants

of the British Isles without exception, were by the beginning of

the Roman period Celtic. Here again, views prevalent not long

ago require modification. The late Sir John Rhys called atten-

tion to a broad division of Celtic languages into two classes,

which he called Goidelic and Brythonic, most easily distin-

guished by the existence of a q sound in the Goidelic where the

Brythonic used a p; and he advanced the theory that there had

been two waves of Celtic immigration into Britain from the

Continent, the first of Goidels or ‘Q-Celts’, the second of

1 Archaeologia, lxxi (1921), p. 161. Sir Arthur goes on to suggest that, since our

evidence is derived from ‘Romano-British settlements’, it does not represent ‘the

real British people of the Roman period’, but people ‘who arrived in England

during the period of the Roman occupation’. I venture to dissent from this judge-

ment, because I think that the inhabitants of Roman villas in Britain, about whom

Sir Arthur is here writing, were in general not immigrants but Romanized natives:

I believe, therefore, that their physical characteristics are an index of those pre-

vailing in Britain when the Romans came.

2 A cautious attempt to establish certain local peculiarities of physique in the

Romano-Britons has been made by Dr. L. H. Dudley Buxton in an important

paper on ‘The Racial Affinities of the Romano-Britons’, J.R.S. xxv (1935), p. 35.

P-CELTS AND Q-CELTS

19

Brythons or ‘P-Celts’. Archaeologists then supposed it their business to discover when these two invasions took place, and what distinct types of civilization they introduced. The task was never satisfactorily accomplished; and to-day archaeologists who are also Celtic philologists no longer regard it as a legitimate problem. It is now held that the original language brought to Britain by Celtic invaders belonged to the P-variety, and that the Q-variety did not reach Ireland by way of Britain at all. In confirmation of this view it is pointed out by students of place-names that, throughout Great Britain, no trace can be found of Goidelic names, except where they can be explained as due to migration from Ireland at a time subsequent to the beginning of our era.1

It is not known when the Celtic language was introduced into Britain; the attempts to correlate that event with any of the various prehistoric invasions known to archaeology have been, so far, conjectural. It would be generally agreed that it cannot have been later than the beginning of the Iron Age, that is to say, roughly, the sixth century before Christ. One thing is clear: it was sufficiently early for pre-Celtic speech to have left no recognizable trace upon the British names of people or places which, from the time of Caesar onwards, Roman authors committed to writing. It is natural to expect that place-names still in use to-day might preserve such traces; but even in the case of river-names, the most conservative of all, no examples can be quoted with certainty. Professor Ekwall2 sums up a discussion of the whole subject by saying: ‘I cannot point to any definite name that strikes me as probably pre-Celtic.’ At the same time it is possible, as some scholars maintain, that certain Celtic dialects in western Britain have been influenced by pre-Celtic speech; though this, if it is true, is no evidence of such speech surviving as a living language into the Roman period.

The material civilization of the Britons in the first century before Christ was by no means uniform. Archaeology, especially within the last twenty years, has told us enough about it to make possible a tentative survey of its characteristics; but this survey must be somewhat like a geological map, where one formation is shown as superimposed upon another, and the

1 I. C. Peate, ‘The Kelts in Britain’, Antiquity, vi (1932), p. 156.

2 English River-Names (1928), pp. liv–lv.

20

BRITAIN IN THE TIME OF JULIUS CAESAR

upper formations stop short at certain lines, beyond which the

lower go forward without them. It will be enough here to

attempt such a survey for that part of Britain which ultimately

became Roman, omitting Ireland and most of Scotland.

The general background of British civilization at this time

may be described as having a Late Bronze Age character; but

over a great part of its area this background had been overlaid

by newer cultural formations derived from the continental Iron

Age. It will be necessary in the sequel to distinguish four of

these; but first the general background must be describecd.

The Late Bronze Age was a period when Britain as a whole

was a backward country by comparison with the Continent;

primitive in its civilization, stagnant and passive in its life, and

receiving most of what progress it enjoyed through invasion and

importation from overseas. Its people lived either in isolated

farms or in hut-villages, situated for the most part on the gravel

of river-banks or the light upland soils such as the chalk downs

or oolite plateaux, which by that time had been to a great

extent cleared of their native scrub; each settlement was sur-

rounded by small fields, tilled either with a foot-plough of the

type still used not long ago by Hebridean crofters, or else at best

with a light ox-drawn plough which scratched the soil without

turning the sod; the dead were burnt and their ashes, preserved

in urns, buried in regular cemeteries. Thus the land was

inhabited by a stable and industrious peasant population,

living by agriculture and the keeping of live stock, augmented

no doubt by hunting and fishing. They made rude pottery

without a wheel, and still used flint for such things as arrow-

heads; but they were visited by itinerant bronze-founders able

to make swords, spears, socketed axes, and many other types

of implement and utensil, such as sickles, carpenter's tools, metal

parts of wheeled vehicles, buckets, and cauldrons. Judging by

the absence of towns and the scarcity of anything like true

fortification, these people were little organized for warfare, and

their political life was simple and undeveloped, though there

was certainly a distinction between rich and poor, since many

kinds of metal objects belonging to the period imply a consider-

able degree of wealth and luxury.

This civilization differed a good deal in different parts of the

country. In particular, towards the north and west, it was

strongly tinged with Neolithic survivals, sometimes so strongly

THE BRONZE AGE TRADITION

21

as to give it an ambiguous character as between Neolithic and Bronze Age types; in these parts inhumation, the Neolithic burial rite, sometimes lingered on, and the megalithic tradition led people to build their huts and villages with large stones, often placed on end; the soil, too, was poorer and the whole standard of life lower.

The chronology of this period is very vague. The Late Bronze Age in southern Britain is regarded as beginning about 1000 B.C. and lasting until about 400 B.C.; elsewhere it is thought to last to about 200, and in Scotland to as late as the beginning of our era.1 These dates, however, are not only mere approximations, but the events which they are designed to fix are only the beginnings of slow and gradual processes. The peasant-civilization described above may have taken shape after 1000, but it did not cease to exist at the date when the Bronze Age is said to have closed. What happened at that date was that it began coming at certain points under the influence of Iron Age fashions of life and thought, which in many respects left it substantially unchanged. From this point of view the earlier phase of the Iron Age may be better regarded as a period of overlap between the Bronze and Iron Ages; a period when some settlements, planted by new-comers from the Continent, showed a predominant Iron Age civilization, while others went on their way little if at all influenced by their new neighbours. In fact, so important is this overlap between the two periods that some archaeologists believe the characteristic implements of the Late Bronze Age not to have been introduced until the first appearance of iron; while it is an established fact that such implements remained in use, even in a fairly advanced region of southern England, until the last century before Christ. When we speak of the Iron Age as beginning at a certain date, therefore, it must be remembered that this does not imply a general abandonment of Bronze Age fashions, nor even the general beginning of such abandonment; and hundreds of years after such a date the majority of Bronze Age villages were still carrying on their old life with little change.

Of the various Iron Age cultures2 which had by the time of

1 These dates are those given in V. Gordon Childe, The Bronze Age (1930), pp. 236–7.

2 I am here in the main following C. F. C. Hawkes (‘Hill-Forts’, in Antiquity, v (1931), pp. 60–97; Kendrick and Hawkes, Archaeology in England and Wales, 1914–31 (1932), ch. x. My first or south-eastern culture is his Iron Age A; my second

22

BRITAIN IN THE TIME OF JULIUS CAESAR

Julius Caesar established themselves upon this Late Bronze Age foundation, the first made its appearance in the seventh century before Christ and developed especially in the sixth.

The La Tène period had not yet begun; consequently the civilization brought in by these new-comers was based upon, though it did not very closely reproduce, the Hallstatt type1 which flourished on the Continent in the first half of the last millennium before our era.

It is in south-eastern England that the remains of this Hallstatt immigration are chiefly found; but it impinged upon a long coast-line, extending from Yorkshire to Dorset, and its earliest appearance seems also to have been its most northerly.

This was the settlement on the Castle Hill of Scarborough, where, in addition to implements of Late Bronze Age types, a single iron object was found, and much pottery of a kind which shows its makers to have been at least under the influence of the Hallstatt tradition.

The settlement is referred to the seventh century. Other landings of the same general kind have been traced on the Thames, in Kent, and in Sussex; and especially on Hengistbury Head in Hampshire.

The continental affinities of the pottery are in part with south-western France and the Pyrenees, in part with Holland and Champagne; but its general distribution in Britain suggests that its main source lies in the latter direction rather than the former, the migration from Champagne coming at the end of the movement, about 500 B.C.

The influence of this Hallstatt immigration was not confined to the coast. Inland sites affected or freshly settled by it are numerous.

The most important is the village at All Cannings Cross, near Devizes, in Wiltshire, dated to the fifth and fourth centuries, where implements of Late Bronze Age types were found associated with a great mass of Hallstatt pottery.

Another is at Park Brow, in Sussex, where the Hallstatt invaders seem to have destroyed a village of Bronze Age round huts and built instead their own rectangular houses. In general, settlements

(south-western) and third (north-eastern) are his Iron Age B; and my fourth (Belgic) is the first phase of his Iron Age C. Since all that is needed for the purposes of this book is a broad and generalized picture, I omit much detail and ignore complications, important for the prehistorian, which have been added to Hawkes’s scheme especially by the excavations now in progress at Maiden Castle in Dorset.

1 Bushe-Fox, in Antiquaries Journal, iv. 225, calls attention to the differences which mark off the Hallstatt material in England from that derived from the continental sites to which the name is properly applied.

HALLSTATT INFLUENCE IN BRITAIN

23

yielding the pottery characteristic of this culture occur all over

the south-east, from Kent to the Cotswolds and the Wash.

Many of these settlements indicate a mode of life not perceptibly

differing from that of their late Bronze Age background: they

are farms or villages, often undefended, lying among their little

fields on river-gravels or light upland soils, mostly cremating

their dead, storing their grain in underground pits and grinding

it with primitive querns, not yet made with the upper stone

revolving upon the lower; keeping oxen, sheep, goats, and

pigs; still using bronze and even flint implements and possess-

ing very little iron, but indicating their date by a change in the

style of their pottery, which, however, is still made without

the wheel.

Other settlements are of a new type, and indicate a change

in the political life of the people. These are hill-forts, which

now, for the first time since the Neolithic age, began to be

constructed in large numbers. The hill-forts of this period,

though varying a good deal in size and shape, are all alike, so

far as we yet know them, in being very simply planned: they

consist of a single rampart, which may be of stone, but was

generally an earthwork revetted and palisaded with timber and

protected by a single ditch; the size of the ramparts is not very

great, and the entrances are simply planned, though excavation

has in some cases revealed traces of wooden guard-rooms. These

were not mere camps of refuge. They were permanently

inhabited towns, containing pit-dwellings where the nature of

the soil allowed it. Nor do they seem to have been the strong-

holds of invaders in a hostile land; rather they appear to have

come into existence by degrees, as the newcomers multiplied

and developed a tribal system and tribal wars, and most of

them are dated to the fourth and third centuries before Christ.

This south-eastern Hallstatt culture, although apart from the

construction of hill-forts it changed the face of the country and

affected its civilization very little, was sufficiently powerful to

resist further movements of population from without. A few

objects of the La Tène period found their way into the region

which it occupied, but very few; and until the Belgic invasion

of the last century before Christ (and to some extent, as we shall

see, even after that event) it remained somewhat stagnant and

kept its original character unimpaired: presumably with a

gradual fading away of the Bronze Age elements, and a parallel

24

BRITAIN IN THE TIME OF JULIUS CAESAR

development of the iron industry in the Weald, whose first

exploitation dates from this time.

The south-west shows an Iron Age culture of a different and

later type. About the opening of the fourth century the tin

trade began bringing to Cornwall influences from the Hallstatt

civilization of northern Spain. The clearest evidence for this

connexion is a series of late Hallstatt brooches, found at

Harlyn Bay and Mount Batten, whose origin is to be sought in

Galicia and northern Portugal, not far from the tin-mines of that

region; but pottery found at Chun Castle, a circular stone fort

in Penwith, as well as the shape of the fort itself, which re-

sembles the citanias of Galicia, reinforces this evidence; and

pottery of the same kind, from many sites in south-western

England, shows that the connexion produced widespread con-

sequences in opening the south-west to an Iron Age civilization.

Before long, however, Spanish influence was replaced in this

part of Britain by Armorican. Presumably the Veneti of the

Morbihan, whom Caesar found holding a monopoly of the

coastwise trade round Cape Finisterre, were already beginning

to assert themselves, and to cut off direct communication

between Britain and Spain. Whatever the reason, the fact is

clear that the main Iron Age culture of south-western Britain,

after an initial phase in which it was dominated by that of

Hallstatt Spain, settled down into a La Tène culture in close

connexion with Brittany.

The main development of this culture took place in Dorset

and Somerset, whence it spread along the Cotswolds into the

midlands. Along the eastern edge of this area it marched with,

and in many places supplanted, the earlier south-eastern culture

already described, being superior to that culture both in its

material equipment and in its political and military organiza-

tion. Evidences of the character of its daily life have been

richly provided by the lake-village at Glastonbury, which proves

it by far the most advanced civilization by that time established

in this country. The pottery is of good quality, skilfully made

though not thrown on the wheel, and the quantities of it very

large. Metal objects, both in bronze and in iron, are numerous,

and metallurgy was practised in the village itself. Numerous

and well-made wooden objects were found, including portions

of looms and carts, as well as tubs, spoons, and other small

things. And not only are all these things skilfully made, but

LA TÈNE IMMIGRATIONS

25

a great many of them are beautifully decorated in that elegant,

flowing, curvilinear style called by English antiquaries Late

Celtic, which now for the first time makes its appearance in

this country.

The political and military power of this civilization is attested

by its hill-forts, which exist in great numbers and far exceed

those of the south-east in the massiveness of their structure and

the intricacy of their design. Their defences ordinarily consist

of stone walls protected by ditches; and multiple ditches, com-

plicated outworks, and additional defences at the gateways are

common features.

Eastward the spread of the new culture was blocked by the

already established Iron Age civilization of the south-east: but

westward it was able to range freely, if somewhat superficially,

over the highland zone, whose Late Bronze Age civilization

was incapable of resisting it; and hence it is found pushing its

way into Wales, establishing hill-forts at Lydney and Llanmelin

in the southern marches and leaving relics of its decorated

metal-work at various places in the centre and north.

Simultaneously with this south-western La Tène culture,

another of the same kind was developing in the north-east.

This is especially connected with the name of the Parisi, a tribe

of the East Riding mentioned by Ptolemy, whose presumable

origin from the Parisii of the Seine valley is confirmed by

excavation of their burials, where warriors were interred with

their weapons and their chariots in a manner recalling the

middle La Tène chariot-burials of the Marne. Here, again,

La Tène art is found in a highly developed form; and from this

source it spreads widely over the north of England, penetrating

into Scotland and even into Ireland. But there is nothing in

this northern culture to set alongside the peaceful village-life

of Glastonbury. We have, rather, an aristocracy of invading

warriors, establishing their dominion over a people whose daily

life is still characteristically Bronze Age. The relics of La Tène

art in the north at this period are scabbards, shields, horse-

trappings, and in general the implements of war; their resem-

blance to things found in France is so close that some of them

were almost certainly brought in by the invaders, to form the

basis of a native school of which more will be said hereafter.

In spite of the contrast between these two La Tène immigra-

tions, the warrior aristocracy of the north-east ruling over a

26

BRITAIN IN THE TIME OF JULIUS CAESAR

native population of alien culture and the richer and more

peaceful civilization of the south-west containing in itself

certain peasant elements of its own, there are also strong

similarities. One is their common dependence upon continental

sources of the middle La Tène period. Another is the fact that

they established themselves on the outer fringe of the area

already occupied by the Iron Age culture of the south-east,

and by degrees joined hands, outside the limits of that area, in

the midlands, so that no definite frontier can be drawn between

them. A third is their common situation on the line dividing the

highland and lowland zones, with the result that in both cases

they pushed outward into the highland zone and there estab-

lished an ascendancy over a native culture of Bronze Age type,

far more tenacious and conservative than the corresponding

peasant cultures of the lowland zone, ready perhaps to accept

new masters but by no means ready to accept new fashions of

life. In the north this feature is illustrated by the numerous

hut-villages of the Brigantian region, wholly Bronze Age in

general character, where excavation only very rarely yields a

fragment of recognizably La Tène work. In the south-west it is

illustrated by such settlements as Chysauster, in Cornwall, whose

chambered houses belong to a type of architecture characteristic

of the outermost fringe of the highland zone, and represented

at its highest development by the brochs of north-western

Scotland. At Chysauster, however, the La Tène objects found

in these houses are far more numerous than those found in any

hut-village of the north; a fact which, with others, goes to show

that the influence of La Tène civilization over the Bronze Age

peoples among whom it was established was greatest in the

south-west and diminished northwards, being in Wales very

slight before the beginning of our era, and in the north of

England practically non-existent.

The last of these Iron Age invasions, and to historians the

best known, because it led directly to the interference of Julius

Caesar in British affairs, was that of the Belgae in the first

century before Christ. For this event we have the testimony

of Cacsar himself. Within the lifetime of his informants, he

says, the Gaulish king Diviciacus had governed Britain as well

as a great part of north-eastern Gaul, and later Commius,

whom he made king of the Atrebates in Gaul, was a man of

influence in Britain; statements whose significance becomes

THE BELGIC INVASION

27

clearer when he tells us that the parts of Britain nearest to the

sea (that is, the straits of Dover) were inhabited not by people

born in the island but by people who had crossed over from

Belgic Gaul, at first to plunder, then to settle and till the soil;

and that almost all of these still bore the names of the Gaulish

tribes from which they were sprung.1 He emphasizes the wealth

and dense population of the region thus settled, and reckons its

inhabitants as the most civilized of the Britons, differing little

in manners and customs from the Gauls.

The evidence of archaeology supports and supplements these

statements. A peculiar culture belonging to the late La Tène

period was identified in the last century in a cemetery at

Aylesford, in Kent, and its characteristic pottery has since then

been found in Britain over an area spreading across the Thames

to Hertfordshire and Essex, and reaching at its greatest exten-

sion to Northamptonshire and the Cherwell valley. In origin

this culture was the product of a fusion between the middle La

Tène civilization of the Marne and Aisne valleys and new settlers

of German blood, arriving in that region in the second century

before Christ; for the closest parallels to the material found in

Britain come from the southern Belgic region, in particular the

territory of the Catalauni, round Châlons-sur-Marne. The date

of the migration is fixed by Caesar's evidence to within a

generation before his own activity in Gaul, that is, to about

75 B.C.; and by the time of Caesar's expedition the Belgic area

in Britain included the east and north of Kent, a few landing-

places on the coast of Essex, and at least a considerable part of

Hertfordshire.

The character of these new-comers' civilization reflects their

mixed pedigree. Hitherto all inhabitants of Britain had at-

tached themselves to what in the last chapter were called areas

of primary settlement: the lighter soils, whether on the uplands

or on the banks and terraces of rivers, which were naturally

drained and unencumbered by heavy timber. For purposes of

defence they had fortified themselves on hill-tops or promontories. The Belgae, on the Celtic side of their ancestry, inherited

these same traditions; but on the German side they had the

instincts of forest-dwellers, accustomed to living in woodland

clearings, and an agricultural technique adapted to working

heavy soils. Even at Glastonbury, the most advanced agricultural

1 Bell. Gall. ii. 4; iv. 21; v. 12.

28

BRITAIN IN THE TIME OF JULIUS CAESAR

implement actually discovered is the foot-plough, and until

the Belgic invasion it is fairly certain that no animal-drawn

plough was used in this country except a light and primitive

type which merely furrowed the surface of the soil; but the

Belgae appear to have brought with them the heavy Gaulish

plough armed with a coulter and designed to turn a sod. Hence,

for the first time in British history, the Belgae began to clear

forests, and to exploit the heavier soils on which they grew; they

covered the country with isolated farm-houses1 and stocked it

with cattle; they settled in the valleys and appropriated river-

crossings; they did indeed construct hill-forts, but much less

systematically than the earlier Iron Age settlers, and in the

course of time-though this only happened after the Julian

invasion-they abandoned the hill-tops and built their towns

on valley-sites which have been inhabited ever since. Even

where, in the early days of their settlement, they made hill-

forts, these were so far influenced by their forest-dwelling habits

as to excite special remark from Caesar, who stormed two of

them. 'The Britons call it a fort', he writes, 'when they have

defended a tract of dense woodland with a rampart and ditch';

a description which could never have been applied to the earlier

hill-forts of the open downland. They also introduced into

Britain a practice which we know to have been current among

the early Germans: that of delimiting their territory by means

of a continuous rampart and ditch. A boundary of this kind,

dated to their time, runs across the Hertfordshire plain from

Verulam to Wheathampstead, with another sector north-west

of Verulam; it was evidently made in the early days when they

were extending their power north of the Thames, that is, just

before the Julian invasion.

Their pottery, made on the wheel, was technically much

superior to anything hitherto made in Britain; but both here

and elsewhere they showed themselves defective in the artistic

qualities with which the earlier La Tène invaders were so

richly endowed. It is not until a later period, when no doubt

their political predominance enabled them to attract and

develop artistic talent, that we find decorative art growing up

among them. At this early date they reveal themselves as good

1 Caesar's creberrima aedificia, I take it, can only be explained in this sense.

The point will become important when we have to consider the origin of the

'Roman villa'.

farmers and good organizers, capable workmen and stout

fighters, men of solid utilitarian temperament: in this, too,

Germanic rather than Celtic in nature.

It was the Belgae who introduced coinage into Britain. The

earliest coins found in this country, apart from stray specimens

of Mediterranean fabric brought by trade to the south-west, are

those of the Gaulish Bellovaci, apparently struck early in the

first century B.C., which are found over an area almost corre-

sponding with that of the Belgic settlements; other types assign-

able to the Atrebates and perhaps to the Morini are found in the

same region, sometimes spreading beyond it; these represent

the coins brought with them by the invaders, with some that

may have arrived by way of trade before the migration began.

After the Belgic settlement coins began to be struck in Britain

itself; possibly this was already happening before Caesar's in-

vasion; but that is not certain, and in any case the main develop-

ment of British coinage belongs to a later date.

In speaking of coinage in Britain, Caesar mentions that iron

bars of standard weight were also used as currency. He seems

here to be referring to certain objects chiefly used not by the

Belgae but by the south-western tribes: objects resembling a

half-finished sword with a roughly formed handle, of which

large numbers have been found in Somerset, Dorset, Wiltshire,

Gloucestershire, and Worcestershire. It has been contended that

they are not in fact currency-bars, but merely unfinished swords;

nevertheless the traditional explanation of them is probably

correct, and if so they represent a form of currency used by the

La Tène tribes of south-western Britain before coinage proper

was introduced into the country at all.

It is time to attempt a generalized sketch of the civilization of

Britain, as these various movements and developments had

moulded it by the time of Julius Caesar.

In east Kent and round the mouth of the Thames a dense

population of vigorous Belgic settlers were scattering their farms

over the country-side, clearing forest, breaking up new land with

their heavy ploughs, and reaping rich crops of wheat off the virgin

soil. They were divided into small tribes often engaged in petty

warfare, but their main concern was to push forward their own

area of settlement towards the north and north-east. To the

south they were hemmed in by the forest of the Weald, in-

habited only by iron-miners belonging to an older civilization.

30

BRITAIN IN THE TIME OF JULIUS CAESAR

This civilization extended from the limits of the Belgic area to the Wash, the forests of the midlands, and the uplands of Wiltshire; it was a mixture of primitive Bronze Age elements with elements akin to those of the Hallstatt culture, to some extent affected by La Tène influences, but on the whole stagnant and backward; a peasant civilization, housed in hut-villages and scratching the lighter soils, its chiefs mostly living in hill-forts on the open downs, and even its larger tribal units—Trinovantes, Iceni, Regni—having no capital that could deserve the name of a city.

West of this lay a richer and more progressive region. The Cotswolds, Somerset, Dorset, and the lands surrounding them, supported a large population of well-organized tribesmen, primitive in their agriculture compared to the Belgae, but hardly less powerful in politics and war, and infinitely more gifted in the arts. North-eastward they were in communication with a culture like their own which dominated the midland forests from the fortress of Hunsbury, and this again linked up with the warriors of Lincolnshire and Yorkshire, who were gradually asserting their supremacy over the hut-villages of the lingering northern Bronze Age and laying the foundations of the Brigantian kingdom.

The westernmost fringe may by now have learned everywhere to talk Celtic, but its civilization was practically untouched by Iron Age influences. Cornwall had its tin trade and its ancient connexions with the civilized world; but Devonshire, Wales, and the north-west were still occupied by a backward and poverty-stricken Bronze Age population, living in hut-farms and hut-villages and owning hardly any implements except of wood and stone.

It sometimes happens that the name by which a nation or country is known to history perpetuates some error on the part of those who first used it, as when the natives of the American continent are called Indians because the European explorers who discovered it were in search of India, or when we call the Hellenes Greeks because the Romans transferred to them the name of a small Epirote tribe. The origin of the name Britain is obscure; but in all probability it enshrines a blunder of a somewhat similar kind.

Among the Belgic tribes inhabiting the continental shore of the straits of Dover in the time of Julius Caesar was one which

THE NAME BRITAIN

31

Pliny1 calls Britanni. From his text, they appear to have lived immediately south of Boulogne. The Belgic invaders of Britain were drawn not from this region, but from a district a good deal farther to the south; but this was the region from which Caesar's expedition sailed, and his port of departure lay almost in the Britannic territory; it may therefore be assumed that he knew their name.

But the country which ever since he invaded it has been known as Britain—Britannia, the land of the Britanni—was not known by that name to earlier explorers. Although manuscript readings are discrepant, no doubt through a tendency to replace the earlier name by the better-known later, they contain evidence that from the fourth century before Christ, when Britain was visited and described by the Greek traveller Pytheas, the British Isles were called by many geographers, Diodorus Siculus, Strabo, Ptolemy, and others, the Pretanic Isles. Scholars agree that the name Pretani, which is thus implied as a general name for their inhabitants from at least the fourth century onwards, can hardly be identical with the name Britanni. On the other hand, there is a name with which it not only can but must be identical: it is the old Welsh Priten, the 'P'-form of which the corresponding 'Q'-form is Cruithin, the name by which Irish writers refer to the Picts. The name Picti, painted men, is a Latin translation of this word, appearing in the third century as a general term by which Latin-speaking Britons referred to the barbarian tribes beyond the frontier; later it was adopted by these tribes themselves, and became the title of the Pictish kingdom whose people were called Cruithin by the Irish.

In the light of these facts it can hardly be doubted that the true form of the name Britannic, as the name of Great Britain and Ireland, is really Pretanic or Pritenic, and that the inhabitants of these islands, perhaps already before the beginning of the Iron Age, were called Pretani or Priteni. It is a very probable conjecture that Caesar, finding himself at the outset of his expedition on the borders of a Belgic tribe called Britanni and knowing that the country he was about to invade had been lately colonized from Belgic Gaul, believed himself able to correct a widespread error by substituting the forms Britanni, Britannia, for Pretani and its Latinized correlative Pretania.

1 Natural History, iv. 31.

III

CAESAR'S INVASION

WHAT motives induced Caesar to attack Britain, what he intended to bring about there by his invasion, and how long the project had been shaping itself in his mind before he set about executing it, are questions to which he has given us no answer. Yet we cannot help asking them; and unless we can find some sort of answer, at least to the first and second, the mere narrative of his campaigns must remain unintelligible.

The nearest approach in Caesar's text to a statement of his motives for the invasion is the remark that, in almost all his Gallic campaigns, contingents from Britain had been fighting on the side of his cnemies.1 In other passages also, he testifies to the close connexions between Britain and Gaul: in 57 B.C. those chiefs of the Bellovaci who had urged resistance to him took refuge in Britain when their policy broke down; he tells us that in his time there were chiefs whose power extended to both sides of the Channel; and in his general sketch of Gallic civilization he describes Britain as the cradle and school of Druidism. Every reader of his Commentaries knows that, as the years of his work in Gaul went on, he became more and more preoccupied with the problem of keeping the peace, and forestalling or repressing movements of rebellion, among the fiercely independent tribes he had conquered. It was the necessity of guarding against such risings that dictated his last acts in Gaul, before he crossed the Channel on his final British adventure; it was fear of their imminence that forced him to abandon it before he had carried it to a finish; there is little room for doubt that the same motives played a part in determining the enterprise itself.

As the event of his expedition showed, Caesar was on the horns of a dilemma. So long as Gaul was restless, Britain, a refuge and reservoir of disaffection within a few hours' sail, was an added danger : for the sake of Gaulish security, therefore, Britain must be made harmless. But so long as the restlessness of Gaul was acute a campaign across the Channel was hazardous:

1 B.G. iv. 20. It should be noted that Caesar gives this as a reason not for invading Britain at all (for that, he gives none) but for pushing the invasion forward as rapidly as possible.

HIS PLANS

it was an incitement to revolt in Gaul while the Roman armies were overseas. Either way there was a risk. Caesar chose one risk and went to Britain: Augustus, warned by his example, chose the other and stayed in Gaul.

The question which risk to take could be decided only in the light of an estimate of possible gains; that is, in the light of an answer to the question what a British campaign might hope to achieve. It is impossible to think that Caesar had no answer of his own to this question; but he has not told us what it was. The last thing he did before sailing to Britain in 55 was to cross the Rhine and undertake a campaign in Germany; and in narrating the events of that campaign he ends with the statement that his plans had been fully carried out and his objects completely achieved. At the end of his British narrative he makes no such claim, and the contrast is significant. His reconnaissance in 55 was followed by a twenty days' thanksgiving at Rome to celebrate the happy commencement of a new adventure. Nothing of the kind followed the definitive campaign of 54. Whatever the object of the expedition may have been, neither Caesar nor the Senate thought that it had been fulfilled.

For this reason it cannot be supposed that Caesar meant only to conduct a punitive war, a large-scale raid, in order to teach the Britons that they were within the reach of Rome's arm, and to warn them against interference in the affairs of Gaul. That is what he meant to do in Germany; had he meant to do only that in Britain, he could have justly claimed that he had done it. He meant to do more.

He had taken up his command in Gaul in 58, with four legions. In that year he did little more than expel the Helvetian and Germanic hordes that threatened it from the east. In 57, with two more legions, he broke the power of the Belgae. In 56 he crushed the Veneti and carried his standards to the Pyrenees and the Bay of Biscay. He had, in fact, overrun the whole of Gaul in three years: by the beginning of 55 'Gaul was, to all appearance, conquered'.1 The appearance was no doubt deceptive: much of the hardest work was still to come; but in estimating his intentions at the beginning of his British campaign we must be guided by the situation as it then appeared, and by his record of achievement in Gaul as it must have stood

1 Rice Holmes, Caesar's Conquest of Gaul, p. 95.

34

CAESAR'S INVASION

in his own eyes at that time. He knew the size of Britain with

a fair degree of accuracy; he knew that its inhabitants were

less civilized and less highly organized both in politics and in

war than the Gauls; he meant in the following year to invade

the country with five legions and to keep them there for the

winter; and when all these facts are considered at once, it can

hardly be doubted that his plan was to conquer the whole

island.

This plan was presumably formed during his Belgic campaign

of 57. It was then that he visited the regions from which Britain

had lately been colonized; and during that summer he had its

existence and importance repeatedly impressed on his mind. It

was then that he heard of Diviciacus, lately king of the Suessiones

and overlord of Belgic Britain; it was then that the hostile

faction of the Bellovaci took flight across the Channel; it must

have been then that he made Commius ruler of the Atrebates

and learnt that he too claimed to be a man of authority on the

other side of the straits. And if Strabo is right in saying that the

Veneti, after submitting to him in 57, rebelled at the beginning

of 56 because they heard that he was about to invade Britain

and feared the loss of their monopoly in the Atlantic coastwise

trade, it follows that Caesar's plan was by then not only formed

but already becoming a matter of common knowledge.

It was late in the summer of 55 when Caesar felt himself free

to turn his attention to Britain. He had destroyed an invading

force from Germany and had completed his punitive expedition

across the Rhine; and he writes as if, after this interruption, he

were picking up the thread of his plans where the German

inroad had compelled him for a time to lay it down. It is clear

too from his words that he expects a critical reader to ask, not,

'Why did Caesar go to Britain at all?' but, 'Why did he go there

with a couple of legions for a hurried visit so late in this year,

instead of waiting for the next summer and then invading in

force?' To this question he gives two answers. First, the con-

stant help which Britain had been in the habit of sending to

his enemies in Gaul: this, evidently, was an inducement to show

his hand and prove to the Britons that their own safety was in

the balance. Secondly, the desirability of a reconnaissance in

order to gain information about the approaches, the harbours,

the country, and the inhabitants.

In view of the close connexion between Britain and the main-

land, a connexion vouched for by Caesar himself, it is natural

to suppose that, to many at least of his questions, answers

might have been given by merchants and others in Gaul. And

of the merchants he did make diligent inquiry, but failed to

discover what he most wanted to know. They were only

acquainted, he found, with the Channel coasts, and even there

they could not tell him of a harbour able to accommodate so

vast a fleet as his expeditionary force would need.

They may have known more than they were willing to tell;

but Caesar himself never hints that they were concealing any-

thing from him; and if he had believed this he had every reason

to state it, as justifying an expedition which, he was well aware,

needed justification. If his plan was to conquer the whole of

Britain, all becomes clear. He asked the merchants, as indeed

his words imply, for information not only about the Channel

coast but about the entire country; they were honestly unable

to give it; and he went over to get it himself. But he could not

afford to explain that to his readers, because in the event he got

no information whatever except about that very coast which

was already well known to his informants. To state the facts

explicitly would have been to confess that his reconnaissance

was a failure. He did, however, see enough to convince him

that what the Gaulish traders had told him was true so far as it

went; and at least he did not attempt to cover his own failure

by accusing them of bad faith.

His preparations were simple. He requisitioned close on a

hundred Gaulish sailing craft to serve as transports, assembled

them at the straits of Dover, and sent a trusted subordinate,

Gaius Volusenus, us to reconnoitre the opposite coast from a

warship. Volusenus was away for the best part of five days, and

in the light of what happened later we can infer that, making

the obvious landfall at the cliffs of Dover, he explored the coast

for some distance to the east and west. From the South Fore-

land to Folkestone he found the beach commanded by ram-

parts of cliff. West of that, the tidal inlet where now is Romney

Marsh gave doubtful access to hilly and wooded country. But

eastward he soon arrived at the open shelving beaches about

Deal, with anchorage in the Downs and plain country, well

cultivated, behind them. Here, he saw, a landing might be

made. Had he gone farther, he would have found what Caesar

wanted: a great land-locked harbour inland from Sandwich,

36

CAESAR'S INVASION

protected by the isle of Thanet, with the low hill of Richborough offering a perfect site for a camp. This was the port used in the following century by the armies of Claudius; and had Caesar known of it the whole course of his British campaign would have been different. Conditions in Gaul would, even so, never have permitted him to complete the conquest of Britain; but his fleet would have been secure, and the chief anxiety that crippled his operations there would not have arisen. The failure to find the port of Richborough was the blunder that marred the entire expedition. And the responsibility for it lies not altogether on the shoulders of Volusenus; Caesar himself, after the hard-learned lesson of 55, looked for another landing-place and came within a mile or two of the harbour without discovering it.

Eighty of Caesar's transports were now in the river-mouth at Boulogne; eighteen more, which he intended for his cavalry, lay a few miles farther north at Ambleteuse. Traders crossing the Channel warned the British tribes of his preparations, and several of them sent promises of submission and undertook to find hostages for their good conduct. Caesar, making use of the tribal connexions between Britain and the Continent, sent Commius back with their envoys to urge as many tribes as possible to submit; and on the return of Volusenus his armada sailed.

A force of about ten thousand men embarked at Boulogne, leaving harbour on the ebb after dark and waiting until midnight to sail. It consisted of the Seventh and Tenth legions, with archers and slingers; the cavalry were given orders to proceed to Ambleteuse, take ship there in the vessels awaiting them, and join the fleet as it passed. It was a fine night in late August; the moon was past the first quarter, and the wind fair, but light and fitful. Towards morning it died away, and the sailing-ships were almost becalmed when, about nine o'clock, Caesar halted the galleys off the little port of Dover, and waited for his transports to close.

The Kentish Britons had meanwhile changed their minds about submitting to Caesar. Perhaps further information from Gaulish traders had revealed the smallness of his force; in any case, when their envoys returned, bringing Commius with them, they found the tribes bent on resistance. Commius was seized and thrown into chains, and preparations were made for war. Charioteers, cavalry, and infantry assembled on the downs above

HIS FIRST EXPEDITION: DEPARTURE

37

Dover, on the assumption that Caesar would probably land there; and consequently, when he drew near the harbour in the morning, he saw both lines of cliff, which overhung the beaches of the inlet to right and left, thronged with armed men.

These were no conditions in which to attempt a landing, even if he had originally intended it; but he knew from Volusenus' report that a better place was to be found farther eastward;1 he therefore anchored while his transports came up on the last of the breeze, summoned his officers to the flagship, told them what he had learned about the lie of the land, and laid before them his plan, urging upon them the necessity of watching for his signals and obeying them smartly: the only way in which he could hope to throw all his force ashore at once and escape being defeated in detail.

The cavalry had ridden to Ambleteuse according to instructions, embarked with unforeseen delay, and failed to clear the harbour, having missed the evening tide. In the morning, when they did sail, they were met by a contrary wind and obliged to put back. Caesar awaited them in vain until the middle of the afternoon, when the tide turned up-channel and a breeze from the south2 sprang up with it. As soon as he saw his ships swing to their anchors, he gave them the signal to weigh and make sail, and the whole fleet stood slowly north-eastward along the coast and rounded the South Foreland.

The Britons divined Caesar's intention, and marched across the downs, cavalry and charioteers leading, to meet him. When he reached the shelving beach selected by Volusenus as the best landing-place, and ran his transports aground on the shingle,

1 I follow Rice Holmes in making Caesar land east of Dover, in spite of Commander H. D. Warburg's attractive argument (Eng. Hist. Rev. xxxiiii (1923), p. 226), for a landing near Hythe. He shows conclusively that if Caesar's dates and times are correct the tide-stream when he weighed anchor was running south-west, and that south-west gales are overwhelmingly commoner, especially at that time of year, than north-east. But he has overlooked the facts that (a) the gale undoubtedly was a north-easterly one, for the cavalry transports running before it were driven westward (iv. 28) ; (b) in 54 Caesar certainly landed east of Dover, and at a place which he had chosen in 55, which therefore cannot have been far from his landing in that year (v. 8). I suspect that something has gone wrong with the numeral quartum in iv. 28; if the interval was more like a week, the stream would be running north-eastwards when he weighed.

2 This was the only wind which would permit the sailing-ships both to round the South Foreland from their anchorage off Dover and to run on shore in the neighbourhood of Deal. It was only a local breeze, or it would have brought up the cavalry from Ambleteuse.

3720.I

D

38

CAESAR'S INVASION

they were ready for him, the infantry a little way up the beach,

the cavalry and chariots at the water's edge. His men, heavily

armed as they were, hesitated to jump over the bows of the

transports into several feet of water and face these strange

engines under a fire of javelins, arrows, and sling-stones; and

it was only when Caesar brought up his warships to reply to

this fusillade, and throw into the balance the heavier fire of

their catapults, that the Britons drew back a few paces and

allowed the legionaries, gallantly led by the standard-bearer of

the Tenth, to gain a foothold on the beach. A confused struggle

followed; but by degrees the legionaries gained ground and at

last formed line and charged. The Britons, however formidable

in a straggling fight, were never able either now or afterwards

to resist Caesar's legions in battle-formation; they broke and

fled, and it was only his lack of horse that prevented Caesar from

completing his victory.

This initial defeat swung the policy of the Kentish chiefs in

a moment to the other extreme. They dismissed their levies,

each man to his farm or village, and bringing Commius with

them came penitently to wait on Caesar, who rated them

soundly and ordered them to send for hostages. For the next

two days the chiefs continued to come in with promises of

submission.

At the end of that time the cavalry transports found the wind

fair though light, and sailed for Britain. The day was far

advanced, and they had nearly reached the camp, when a gale

sprang up from the north-east and drove them out to sea again.

Some, with the wind on the port quarter, reached Ambleteuse;

the rest were blown down channel. They tried to anchor, but

the waves broke over them, and they were forced to stand out

to sea and feel their way back in the dark to the coast of Gaul.

To the credit of their seamanship, not a vessel was lost. These

Gaulish seamen knew their work; but that mystery was not

possessed by Caesar. He had anchored the main body of his

transports in the Downs, and beached his light warships on the

shingle. That night the moon was full, and the tide almost at

springs; and it is difficult to believe Caesar's statement that no

one in his camp knew of the connexion between the two. The

rollers, racing up the shingle before the gale, filled the hulls of his

beached galleys; some of the transports dragged their anchors

or carried away their cables and were dashed on shore under

HIS FIRST EXPEDITION: RETURN

39

the eyes of the helpless Romans. In the morning the beach was strewn with shattered hulks and wreckage. There were no facilities for repairs, no spare tackle to replace what was lost, and no more ships within call. And Caesar, intending only a brief stay in Britain, had brought no supplies to maintain his men for the winter. The British chiefs in his camp saw their opportunity. In hurried whispers they exchanged an oath to raise the tribes once more, prevent the Romans from replenishing their stores from the standing wheat-fields in the neighbourhood, starve them out, annihilate them, and be rid of Rome for good and all. One by one, they slunk away from the camp.

To a less determined man than Caesar the situation might have appeared desperate. Cast away on a hostile shore with a small force of hungry men, with no hope whatever of relief by his friends, there was only one thing for him to do. The year had been dry and there was a good wheat-crop, standing ripe for harvest. He detailed fatigue parties to reap and bring into camp every day as much as they could. At the same time he turned his legionary artificers to the task of demolishing the worst damaged of his ships and using the materials to repair the rest. It was a race against time. Every day the weather was likely to worsen for a voyage, wheat would be harder to get, and the Britons might be expected to accumulate forces against him. One day their cavalry and chariots lay in ambush where they expected his men to forage, surprised the Seventh legion at work, and were pressing it hard when Caesar, thanks to the watchfulness of his sentrics, was able to bring the rest of his force to the rescue. A spell of bad weather followed, after which the Britons thought the time had come for a decisive blow. They approached the camp and offered battle. But they could not face the legions; they fled, and the pursuing Romans ravaged the country far and wide before returning to camp.

Once more the Britons swung from confidence to despair. Before nightfall they were suing for peace, and Caesar was demanding twice the former toll of hostages. But he knew that he would not be there to receive them. His ships were ready; the wind was fair; the equinox was at hand, and further delay would be folly. That night he sailed back to Boulogne; and of all the tribes that had promised hostages only two sent them.

It is clear from the apologetic tone which more than once creeps into Caesar's narrative of this expedition that his

40

CAESAR'S INVASION

conscience was not at ease. He expected to be criticized for under-

taking it, on so small a scale and so late in the season, at all.

He expected to be criticized for making no provision to ensure

the safety of his fleet in a gale at high water of springs. He

expected to be criticized for not having brought supplies, both

in the way of foodstuffs and of materials for repairing his ships

in the event of damage. None of these criticisms can be alto-

gether rebutted. There are only two things that could be said

in Caesar's defence, and neither is conclusive.

One is that he did not know how dangerous a matter it was

to ship an army to Britain without having first secured a

sheltered harbour for his transports. Had he understood the

conditions and the perils of ocean as opposed to Mediterranean

navigation, he would either never have gone to Britain in the

late summer of 55, or he would have insisted on a sheltered

harbour or a reserve fleet, or both. Yet it is necessary to record

the fact that even his misfortunes in that year failed to teach

him the required lesson, and he repeated his mistake in the year

following.

The other plea in his defence is that he did after all succeed

in the main object of his expedition. It was not an invasion, it

was a reconnaissance; and he came away from Britain having

learnt something about the country and the people. But what

did he actually learn? As to how a fleet could go safely to

Britain and remain there safely until it was wanted again he

learnt nothing. He learnt nothing about the geography of

Britain as a whole, its natural resources, or its military power.

He only came into contact with a handful of East Kentish

tribes, and concerning these, it is true, he gathered certain

useful information, though none, surely, which the Gaulish

traders had not already given him.

He learnt that the Britons, unlike the Gauls of his own time,

made use of chariots in warfare, and he saw how these were

handled. They served partly as a kind of cavalry, partly as a

kind of mounted infantry used in conjunction with cavalry.

The chariot carried a highly skilled and well-armed warrior,

who might either travel independently about the battlefield dis-

charging arrows and sling-stones, or else combine with the

cavalry, join in their charges, and on reaching their objective

dismount and fight on foot while his chariot waited near by to

remove him at need elsewhere. Caesar describes these tactics

HIS FIRST EXPEDITION: RESULTS

with care, and pays a high tribute both to the skill with which

the chariots were driven and fought and to their effectiveness

in battle. At the same time, it appears from his narrative that

the combined cavalry and chariot attack which he describes,

though alarming to his men by its novelty and effective when

used against scattered parties of infantry engaged in foraging

or ravaging, was not dangerous to a legion in battle-formation.

The chief tactical problem presented to him by the British

charioteers was the difficulty of pursuit. On his reconnaissance

in 55 Caesar was totally unprovided with cavalry except for the

thirty men of Commius' bodyguard. Even in his invasion of

the following year he was, like all Roman commanders of the

period, inadequately furnished in that arm. Consequently

charioteers could always escape him when their attack failed.

It may be added that although a few ancient writers describe

the British war-chariot as having its axle armed with a scythe

Caesar neither says nor implies anything of the kind; his

account of their operations rather implies that they had none,

and this inference is borne out by the silence of the best ancient

writers and by the remains actually found in Celtic chariot-

graves, whether in Britain or in Gaul.

Caesar also learnt that in at least the south-eastern part of

Britain there was abundance of grain to be had at harvest-

time; but his provisions for the invasion of 54 show that he did

not rely on this for the support of his troops. Another lesson

was that the British tribes could easily be brought to sue for

peace by sustaining a defeat, and this, we shall see, was an idea

which helped to control his initial strategy in the follòwing year,

with disastrous results.

Taking all this together, one cannot believe either that the

information gathered in the reconnaissance came up to Caesar's

expectations or that it was worth winning at the expense of so

considerable an effort and at so grave a risk. Had the gale been

worse, so as to destroy a majority of his ships instead of a

minority; had the Britons been more energetic in preventing his

troops from foraging; above all, had they collected a fleet

powerful enough to hamper the movements of his returning

transports, which they might well have been able to do for all

he seems to have known about them, Caesar's career would

have ended ingloriously on the coast of Kent. Yet, after all this

is said, the judgement of history depends not on what might

42

CAESAR'S INVASION

have happened but on what did happen; and the fact is that Caesar, like many another great commander, took risks, partly with his eyes open and partly blindfold, and that by the coolness and soundness of judgement which he invariably displayed in an emergency he extricated himself, not on this occasion only, from dangers which to a lesser man would have proved fatal.

It was late in the autumn before Caesar left the winter-quarters where he had placed his legions in Belgic Gaul. In the meantime he had learnt from the fewness of the hostages sent across the Channel that the Britons as a whole did not mean to submit easily, and he had worked out elaborate plans for next year's campaign. He designed a new type of transport, instead of contenting himself with the standard pattern of Gaulish ship: they were lower in the freeboard and broader in the beam, and fitted with oars as well as sails. He arranged for the building of six hundred, to be done by military labour in various parts of northern Gaul: sixty of them, we know, were to be built on the banks of the Marne; rope was ordered from Spain, and the whole fleet was to be ready in the spring.

In June 54 he was back in the north, to find his shipwrights well forward with their work; but the shadow of approaching troubles in Gaul already overhung his preparations: the Treveri were planning revolt, and many of the Gaulish chiefs were lukewarm or worse. But Caesar was not to be turned from his project. He patched up the affairs of the Treveri as best he could in the time, and resolved to take with him the most discontented of the Gaulish leaders, to be hostages for the good behaviour of their tribes.

His force was now ready to sail. Eight legions were encamped at Boulogne, beside slingers and archers and four thousand Gallic horse; but of this army more than a third was to be left behind, in command of Caesar's ablest officer, Labienus, not only to protect the base of the expedition and to keep it supplied with provisions, but also to keep the peace in Gaul. His fleet consisted of 28 warships and 540 transports; 60 more were weatherbound at the mouth of the Seine, but he could do without them; and in addition there were over 200 privately owned vessels belonging to merchants and adventurers who had attached themselves to the expedition.

For over three weeks the armada lay in port, waiting for a

SECOND EXPEDITION: THE START

43

favourable wind. It was an anxious time for Caesar. Among

the Gaulish notables in his camp were some whose disaffection

only awaited opportunity or incitement to show itself; and if it

was his business to see that opportunity was lacking, incitement

was not. Their ringleader was Dumnorix the Aeduan, who had

openly declared in his tribal council that Caesar meant to make

him king, and had never forgiven him for falsifying the boast.

During these weeks Dumnorix tried by every means in his

power to persuade Caesar to leave him behind, pleading his

terror of the sea and his religious obligation not to set foot on

shipboard; finding this vain, he turned his attention to per-

suading his fellow chiefs of the danger they ran by putting their

lives in Caesar's hand; and at last, when all else had failed and

the wind was fair for crossing, he summoned the cavalry under

his command and, in the confusion of a general embarcation,

fled from the camp.

Not even the long-delayed fair wind could distract Caesar's

mind from the peril of this new emergency. He cancelled his

orders for embarkation and sent a strong body of horse in

pursuit of Dumnorix, to bring him back dead or alive. But

Dumnorix had gone too far to retreat; he refused to come back,

and was cut down, crying out that he was a free man and

member of a free tribe. His ghost had its vengeance. It was

the Gaulish freedom which he invoked with his last breath that

ruined Caesar's plans for the conquest of Britain.

That evening the fleet sailed. Towards midnight the light

south-west breeze died away and at dawn Caesar found him-

self drifting up-channel on the flood tide, and saw the South

Foreland receding on the port quarter. But his transports were

now provided with oars, and when the tide turned he set his men

to row, and work their way across the stream of the ebb to the

landing-place he had chosen the year before. By midday all

the ships were beached, and the army landed unopposed.1

The exact position of this second landing is uncertain. Caesar's

words may imply either that it was identical with the landing-

1 This time there is no difficulty about the tides. Caesar left Boulogne on the

ebb, shortly before sunset, which happened about 8 p.m. The tide turned about

10, and the wind failed at midnight; between 3 and 4 a.m. it was light enough to

see the coast, and by that time that was carrying the fleet towards the

North Sea had already begun to slacken. By about 4 they had got their oars out,

and soon after this the tide definitely turned. They must have been then about

ten miles east of the South Foreland.

44

CAESAR'S INVASION

place of the previous year, which, if his own estimate of seven

miles from Dover is correct, must have been close to Walmer

Castle, or that it had been chosen by himself in that year for

future use, in which case it cannot have been far away. It was

not so far north as Sandwich, for had Caesar landed there he

would have discovered the harbour of Richborough and his

ships would have been safe. As it was, he allowed political and

military considerations to determine his next step. Natives

were brought in, who said that the tribes had at first meant to

oppose his landing as before, but on seeing the vast size of his

fleet had changed their minds and retreated into the interior.

Caesar knew the value of a swift and unexpected stroke; he

had learnt in 55 that the Britons had no stomach for a defeat,

and he was confident of his power to defeat them whenever he

could bring them to battle; he decided therefore that this was

an opportunity not to be lost, and determined on a night march

to be followed by a battle at dawn.

Leaving his fleet at anchor and a sufficient force of men to

protect the camp, he marched inland with his main body soon

after midnight. Early in the morning he was looking down into

the Stour valley from the hills east of Canterbury, with the

low sun behind him, and descried the British forces moving

down from the opposite woods, cavalry and chariots leading, to

dispute his passage of the river. He launched his own horse

against them and they withdrew into the woods.

There is an ancient fort, now called Bigbury, on the brow of

the hill two miles west of Canterbury, overlooking the river.

It had been constructed as a fortified town by the Belgae not

long before Caesar's time, and their weapons, agricultural

implements, and household goods have been found there. Its

defences were simple: a bank and ditch, doubtless a palisade,

and an abatis of felled trees in the gateways. Here the Britons,

demoralized by the shock of the first encounter, took refuge.

Woods grew around it, and among these their skirmishers were

moving in expectation of the assault. Caesar swept them aside

with his cavalry and threw the Seventh legion against the

fortress. The rampart and palisade together rose some twenty

feet above the ditch-bottom, but under the locked shields of the

front rank their comrades piled a mound of earth, swarmed over

the palisade, and cleared the place with trifling casualties. No

pursuit was allowed; the men had marched and fought since

THE STORMING OF BIGBURY

45

midnight, the country was strange, and time was needed to

pitch a camp.

If Caesar was satisfied with his day's work, he was much

deceived. Once more the apologetic note makes itself felt in

his narrative when he describes himself leaving his fleet at

anchor 'off a gentle open beach'. He had made a mistake for

which he paid dearly. While he lay encamped in the Stour

valley a gale got up from the east. His ships dragged their

anchors, fell foul of one another, and were thrown by scores on

the beach to be hammered by the full force of the rollers. The

crews left on board could do nothing, and by morning very few

of the vessels were undamaged.

The news reached Caesar just in time for him to stop the

pursuit. Hastening back to the coast, he found that forty of his

ships were a total loss, and the rest not to be repaired without

much labour. He drafted gangs of skilled legionaries to the

work, and sent orders to Labienus to dispatch shipwrights and

begin at once building new vessels to replace those that had

been wrecked. But it was not enough to repair his fleet; he now

at last realized that he must protect it from like damage in the

future. The only way in which he could do this, in the absence

of a safe harbour, was by beaching every ship above high-water

mark and then entrenching them against enemy attack. This

took ten days, which could ill be spared from the short cam-

paigning season; but it had to be done, and Caesar did it.

The negligence with which he exposed himself to a repetition

of the same disaster which he had suffered in 55 is difficult to

understand. Rice Holmes has argued that whereas in 55 his

transports were anchored in the Downs off Walmer, in 54 his

anchorage was in the Small Downs off Sandwich, where, as the

Channel Pilot has it, 'the anchorage is much more secure than in

the Downs, being more sheltered, with better holding ground and

'shoaler water', and where vessels driven ashore on the sand would

suffer less than on the shingle beaches about Deal. This may be

true; and if it were, the contrast between the two anchorages

would have been some excuse for Caesar's conduct in 54; it is

strange, therefore, that he has not mentioned it. All he has urged

in his own defence is that he anchored off a gentle open shore, in

words almost identically repeating the description of the beach

where his fleet was wrecked in the previous year. The only point

of contrast would seem to be that whereas in 55 his warships were

46

CAESAR'S INVASION

beached, in 54 they were anchored with the transports. We can do no more than note the facts that, in spite of his earlier experience, he allowed his impatience to close with the Britons to prevent him from attending first to the security of his ships, and that his intelligence service failed, even after the second disaster, to reveal the existence of a safe harbour at Richborough.1

The earlier mishap to Caesar's shipping had encouraged the Britons to renew their resistance to him; the second had the same consequence, the more so because his armament in this year was on a much larger scale and yet proved equally vulnerable to the hostility of the stars in their courses.

The most powerful man in Britain at the moment was Cassivellaunus, king of the chief tribe among the recent Belgic settlers, the Catuvellauni. This tribe was the spear-point of the Belgic invasion. By now it had pushed its way across the Thames above London and was hard at work consolidating its position in Hertfordshire. Its capital, the stronghold of Cassivellaunus himself, was a fortress a hundred acres or more in extent, protected by earthen ramparts and a ditch still as much as forty feet in vertical depth, that can be traced on the hill-top beside Wheathamstead, a little north of St. Albans. The defences here, and the occupation within them, date from the first half of the last century before Christ; and its excavator says of the fortress: 'It is the work of men with wealth, power, arrogance; but it is the work also of men who were still, for one reason or another, uneasy in their adopted land.'2

When Caesar's threat to Britain became matter of common talk, it found Cassivellaunus fighting to maintain and extend his lately established lordship north of the Thames. Especially he was at war with his eastern neighbours, the non-Belgic

1 Rice Holmes (Ancient Britain, p. 664, note 7) does not face this fact. He says, when refuting some one's suggestion that Caesar may have landed at Richborough in 55, 'it might have been dangerous to land in a harbour with a narrow entrance in the presence of an enemy'; but in 54, even granted that for this reason he might have wished to land on an open beach in the first instance, there could be no reason why immediately on landing he should not have sent his ships round there, a mere two or three miles on Rice Holmes's own theory as to the landing-place, and still less reason why he should not have moved them there after the storm instead of spending ten precious days in beaching them. Richborough was a perfectly good harbour in the time of Claudius; it cannot have been a bad one in the time of Caesar.

2 R. E. M. Wheeler, 'Belgic Cities of Britain', in Antiquity, vii (1933), p. 30.

CASSIVELLAUNUS ENTERS THE WAR

47

Trinovantes. Not long before Caesar’s invasion the Trinovantian

king had fallen in battle, and his young son Mandubracius,

barely escaping with his life, had crossed the Channel and sought

protection from Caesar. With Caesar he returned to Britain,

hoping with his help to recover his father’s kingdom. Though

hard pressed by Cassivellaunus, however, the Trinovantes had

not yet fallen under his rule; nor was he in any sense ruler even

over the Belgic tribes of Britain, for these, as Caesar tells us, had

each a king of its own and were indeed in a state of more or less

constant mutual warfare. Those of Kent made common cause

to resist the Roman invaders, but Cassivellaunus was not a

member of that coalition.

After the fall of Bigbury, which was undoubtedly the capital

of one of these Kentish tribes and, so far as we know, the most

important Belgic city south of the Thames, the Kentish chiefs

reviewed their situation. They saw that there was no hope of

defeating Caesar unaided; and yet, after the events of 55 and

his hurried retreat from the Stour valley when his fleet was

smitten by the gale, they took heart to believe him not in-

vincible. Their one hope lay in enlisting the aid of the great

Cassivellaunus; and the only terms on which they could hope

to do this was by offering him the supreme command.

All operations being at a standstill while the Romans were

beaching and stockading their ships, there was time for the

Britons to arrive at this decision and carry it out before Caesar

was ready to march once more. When he did, the whole

military situation had changed. Cassivellaunus had assumed

control of all the British forces, augmented as they now were

by the addition of his own troops; and Caesar was no sooner

in touch with his enemy than he felt, as any competent general

can feel on such occasions, the personality of their new com-

mander.

He had marched only a few miles, retracing his steps in the

direction of Canterbury, where ten days earlier he had never

met an enemy until his triumphant dash across the Stour and

through the woods of Bigbury, when he found himself engaged

in a running fight with British cavalry and charioteers. He

brushed them aside again and again, but they stuck to their

task and, when driven into the woods, turned against their

pursuers and inflicted heavy losses on them. At the end of a

difficult day’s march, when the Romans were fortifying their

48

CAESAR'S INVASION

camp for the night, the Britons suddenly burst out of the woods and were upon them, hotly engaging the covering troops that had been posted to protect the fatigue-parties. Caesar rushed up the two first and strongest cohorts of two legions in support; for a moment it seemed that the Britons were cut off and surrounded; but there was a gap between the two cohorts, and the British chariots and horse streamed through while the legionaries were too much unnerved to close it, and made good their escape.

It was Caesar's first experience of what British troops could do when well led. The engagement took place just outside the half-finished camp, and every detail of it was visible. It taught him that his heavy-armed infantry, though safe enough in battle-formation, could never put up more than a passive defence against these audacious and swiftly moving barbarians, and that if he tried to pursue them with his horse the charioteers, turning upon them and dismounting to fight on foot, were more than a match for his Gaulish cavalry. They had no regular formation; they fought in isolated and mobile groups, each of which could be reinforced at need by fresh men, or dissolve at any moment into a feigned retreat, to lead pursuers on to their destruction. At the same time, he saw how these tactics could be met, and next day he put his plan into operation. He sent out nearly half his infantry, about noon, to forage: the Britons fell into the trap and attacked them recklessly. But Caesar had kept the rest of his legions and his entire cavalry force, some 1,700 men, in reserve; and the cavalry, gaining confidence from the support of the legions behind them, in a single charge swept the Britons from the field, kept them on the run so that the charioteers had no time to dismount, and routed them decisively. From that day the British levies began to melt away, and never again encountered the Romans in full strength.

By now Caesar knew that Cassivellaunus was the man against whom he was matched, and resolved to end the war by striking at his territory. This meant crossing the Thames; and from the fact that Caesar describes the kingdom of Cassivellaunus as lying eighty miles from the sea it is probable that the ford in general use—certainly not, as Caesar thought, the only practicable one on the lower reaches of the river—was somewhere in the neighbourhood of Brentford. He found it strongly held and obstructed by sharp stakes both above and below the water,

OPERATIONS NORTH OF THE THAMES

49

but the legionaries, whose confidence was now fully restored,

crossed it neck-deep while the cavalry turned the position by

swimming, and the Britons fled.

This brought him into the territory of Cassivellaunus. The

British king, realizing that he could not defeat the Romans in

a general action, had disbanded his infantry, but had no inten-

tion of giving up the struggle. He still had 4,000 chariots, and

these he used with great skill and determination, clearing the

population away from before the advancing legions so that no

prisoners should be taken, and incessantly harassing the Gaulish

cavalry when Caesar sent them out to forage and to burn the

farms. So successful was he that the cavalry could no longer

venture out of touch with the main body, and the work both

of foraging and of destruction was limited to what the legionaries

could do in the course of their day's march.

Pressing thus grimly forward through a deserted country,

whose woods were infested on every side by watchful enemies,

Caesar met a body of envoys from the Trinovantes, welcoming

him as their deliverer from Cassivellaunus, offering him the

submission of their tribe, and promising to accept Mandubracius

at his hands as their king. Their overtures came at the right

moment, for Cassivellaunus had so hampered the Romans in

their foraging that the provisions which Caesar had expected to

find on the march had not been forthcoming. In accepting,

their offers and demanding hostages, therefore, he added a

demand for grain, and marched into the Trinovantian territory

as into a friendly land, ordering his men to respect the inhabi-

tants and their property.

Here he received the submission of five neighbouring tribes,

whose names are otherwise unknown. He also learnt the where-

abouts of the Catuvellaunian capital: and nothing more plainly

shows the ability of his chief enemy than the fact that he had

not learnt it before. Had Cassivellaunus allowed even a handful

of his subjects to be taken prisoner, the secret could hardly have

been kept so long. But now Caesar's next step lay clear before

him. Retracing his path westwards to the marshy valley of the

Lea, he fell upon the Wheathampstead fortress from two sides

at once; the resistance of its defenders was quickly overcome,

and the place fell into his hands with heavy casualties and a

great capture of live stock and prisoners.

But Cassivellaunus had not yet shot his last bolt. While his

50

CAESAR'S INVASION

charioteers were dogging Caesar's march through the Hertford-shire woodlands, he had sent instructions to the four Kentish kings to organize an attack with every man at their disposal upon the naval camp. It speaks well both for his own military capacity that he conceived the plan and for the loyalty of those who had accepted his leadership that, in so dark an hour for himself, they carried it out. Caesar is content to tell us that it was beaten off by a successful sortie in which the defenders suffered little loss; but we learn from Cicero's correspondence a fact which Caesar in his own narrative refrains from mention-ing: that about this time (5 August) he was himself present at the camp. Whether he hurried there because he heard that the attack was imminent, or whether he went when he heard that it had taken place, to see for himself what its effect had been and whether his base was in any further danger, we do not know. In either case the fact of his visit, and the added fact of his silence about it, suggest that he may have been more anxious about the safety of his fleet than he cared to admit; unless, in-deed, as is more probable, it was the state of Gaul that caused him anxiety, and he returned to the Channel in order more quickly to exchange letters about it with Labienus.

Cassivellaunus knew that he was beaten. His only friends in Britain were those of his own Belgic race, and although these had supported him with magnificent loyalty they were now without exception defeated. His capital was destroyed, and, most important of all, his methods of warfare had proved decisively inferior to those of the invaders. The campaign between Caesar and the Belgic Britons had been fought to a finish, and the victory lay with Caesar. There was nothing to do but to make peace. Caesar, it would appear, recognized the inevita-bility of this step on the part of Cassivellaunus, and sent Com-mius to him to expedite it, perhaps by giving him a hint that easy terms would be offered.

Historians have wondered why it was through Commius that the British king approached Cacsar, and have rightly suspected that it could only be owing to Caesar's own initiative. Caesar himself has told us that his reason for making peace with Cas-sivellaunus as and when he did was the fact that he had aban-doned his original intention of wintering in Britain 'because of sudden disturbances in Gaul'. In other words, these distur-bances underlay not only his hasty visit to the coast but also his

SUBMISSION OF CASSIVELLAUNUS

51

employment of Commius as a go-between. If, as we know, he

did not wish to confess that he had made that journey, he would

wish even less to admit that the first overtures came not from

Cassivellaunus but from himself; yet such must have been the

fact. The trouble in Gaul, which was to culminate in the rising

of Vercingetorix, already before the end of his British campaign

determined him to winter south of the Channel. Whether he

suspected that he would never cross the Straits again we cannot

say, but it is clear from his own narrative that henceforth his

only anxiety concerning Britain, so far as this year's campaign

was concerned, was to leave it with a reasonable show of success.

He received the envoys sent by Cassivellaunus and told them

the terms of peace which they were to carry back to their king.

He was to send hostages, pay a stated annual tribute to Rome,

and refrain from molesting Mandubracius and the Trinovantes.

Caesar was still north of the Thames when these negotiations

took place. On receiving the hostages, he marched back to the

Channel and found his transports ready for sea. But they were

not enough to carry his army and all the prisoners he had taken,

whose sale as slaves was to pay the costs of the campaign, and

he decided to make the crossing in two trips. In returning after

the first, he hoped that the empty transports would be accom-

panied by sixty new ships built to his orders by Labienus, but

the wind was foul and only a small part of the expected vessels

reached Kent. For some time Caesar awaited the others; then,

his anxiety to be gone increasing as the equinox approached, he

crowded all that was left of his army and prisoners in the few

ships he had, and on a calm night made his last crossing in

safety.

From a military point of view Caesar was entitled to regard

his invasion of Britain as a success. The chief problem of

strategy had been the discovery and destruction of the fortresses

belonging to the British tribes against which he was fighting;

and this had been done. The chief problem of tactics had been

how to deal with the British charioteers, and this had been

solved not only by the discovery that they were helpless against

a legion in battle-formation, but by the further discovery that

a sufficient body of Gaulish horse, properly supported by in-

fantry, could break them up and rout them by a well-timed

charge. Moreover, in spite of Caesar's failure to discover a safe

harbour, he had learnt how his ships could be protected against

52

CAESAR'S INVASION

the weather; he also knew that wheat in very considerable

quantities could be had in the country; he was therefore right

in thinking that it would be possible for a Roman army to

winter there. Britain was very far from being conquered; but

the first steps towards conquering it had been taken.

From a political point of view the success, if less complete,

was no less real. Over one considerable tribe a nominee of

Caesar's was ruling. Several others had submitted to him with-

out striking a blow. Of the rest with which he had come into

contact, every one submitted before the end of the summer.

Doubtless, as his experience in Gaul had already taught him,

these submissions were precarious and were always liable to be

followed by revolt; but at any rate the political consequences

of his actions in Britain were as favourable as those which he

had achieved in Gaul. On this side also, the omens for eventual

conquest were favourable.

From a financial point of view success was more doubtful.

Caesar knew nothing of the argentiferous lead-mines that were

to play so important a part in the future history of Britain;

indeed, not long before the final expedition sailed, the situation

was well summed up by Cicero, writing to a friend: 'It is now

known that there is not a pennyweight of silver in the whole

island, and no hope of plunder except in the form of slaves.'

The tin and iron of whose existence Caesar was aware offered

no prize to a conqueror. Slaves, however, were to be had in

plenty. The upshot was that Britain might pay for its own

conquest in prisoners of war, but might not prove a very lucra-

tive permanent addition to the provinces of the Empire.

None of these considerations was likely to make Caesar, at

the close of his first campaign, abandon his design of conquering

Britain. There is another question which might have influenced

him in that direction. Throughout his Gaulish command he

knew that he was working against time. It was not enough,

therefore, to believe that he was able to conquer Britain; he

must also believe that he could conquer it fast enough to com-

plete his programme in the time at his disposal. Naturally, he

has told us nothing about this question; we have to ask our-

selves whether, as compared with his record in Gaul down to

that time, the record of his first British campaign was such as

to make him think his progress sufficiently rapid.

The most probable answer is in the affirmative. He believed

A FRUSTRATED PLAN OF CONQUEST

53

that the tribes which he overcame in 54 were the most civilized

in the island, and must therefore have expected the rest to offer

a less powerful resistance. He knew nothing of the difficulties

attendant on a war of conquest in the highland zone, and nothing

of the huge hill-fortresses that guarded the Wessex Downs. Nor,

indeed, are we entitled to say that a soldier of Caesar's calibre

would have been baulked by these obstacles. On the assump-

tion, therefore, that Caesar's original plan was the complete

conquest of Britain, there was nothing in the lessons of the first

campaign to make him change his mind in the autumn of 54.

The fact that he did eventually change his mind arose from

other causes. His conquest of Gaul divides itself into two main

phases. In the first he was receiving the voluntary submission

of some tribes and beating others into submission by pitched

battles; in the second he was putting down determined and

well-organized rebellions among these same tribes by means of

exponsive and protracted siege-operations. His British cam-

paign falls at the line of division between the two. He went to

Britain in pursuance of a plan based on the assumption that the

first phase would prove permanent. It was while he was in

Britain that it came to an end. That very winter his legionary

camps suffered a series of attacks which were the beginning of

the new phase; its ending is haunted by the tragic names of

Alesia and Uxellodunum.

This change in the spirit of Gaulish affairs was the reason

why Caesar's British war broke off at the end of its first cam-

paign. Even at the beginning of the campaign there were signs

in the air of its coming, and Caesar was not blind to them: he

tried to guard against it by taking numbers of Gaulish chiefs

with him across the Channel, and the fate of Dumnorix showed

him that his precautions were not idle. Caesar was not only a

great general, he was a great statesman, but a statesman's ser-

vices to statecraft are to be measured partly indeed by his

successes, but partly by the lessons he teaches through his

failures. One of Caesar's failures was the mistake for which he

paid in the senate-house on the Ides of March. Another was

the mistake which he made when he set out to conquer Britain

before he had pacified Gaul. And in both cases the lesson which

he taught was well learnt by the subtle and patient man who

took up the burden of his inheritance.

3720.I

E

IV

FROM CAESAR TO CLAUDIUS

Caesar’s attempt to conquer Britain had failed; but its failure was due to other causes than the strength of British resistance. With a peaceful Gaul behind it, the attempt, so far as well-informed contemporaries could judge, would have been a success. This, then, was the most obvious double lesson to be learnt both by Britons and by Romans from Caesar’s invasion: that whatever difficulties might arise from the dangers of ocean navigation and the power of British rulers, these difficulties were not insuperable; Rome could conquer Britain whenever she pleased, but not until Gaul was pacified and able to provide a safe starting-point for this further advance.

From Caesar’s own point of view it can hardly be doubted that the conquest of Britain appeared simply as a project deferred, and that the continued freedom of the British tribes was not a reprieve but a respite. We shall find this assumption underlying Roman policy throughout the reign of his successor. The British point of view is less easy to define. A truly statesmanlike British king might have divined and shared the Roman view of the situation; but it would have been natural to take a shorter view, to insist on the fact that an attempt at conquest had been made and had ended in failure, to over-estimate the part played in this failure by the dangers of the Channel and the prowess of British warriors, and to mistake the respite for a reprieve. This at least the Britons could see: that for the immediate future they need not fear another attempt.

Cassivellaunus had sent hostages to Caesar, and this implies acceptance of his terms, namely, that tribute should be paid and the Trinovantes left unmolested. Whether he ever actually sent the tribute we are not told, but it appears that he did refrain from molesting the Trinovantes, and, that being so, the probability is that the tribute was, for a time at least, paid. For a generation and more the Catuvellauni contented themselves with the development of their own territory and its extension northward and north-westward. They no longer needed to protect their capital with defences like those surrounding the Wheathampstead fortress; if they had renounced their eastward offensive at Roman dictation, at least they feared no attack from

that quarter; and it was not very long after Caesar's invasion,

perhaps during the last twenty-five years of the pre-Christian

era, that they abandoned their old capital and built a new un-

fortified one at Verulamium, in the woods overlooking the later

Roman city.

The new capital reflects a new policy. The valley of the Ver

gives easy access into the Chilterns and so, by way of Dunstable,

to the plains beyond them. This was the line taken later by the

Roman Watling Street; and along this line the Catuvellauni

could penetrate without opposition into an almost uninhabited

country, the upper basin of the Ouse. Here we will leave them

for the present, consolidating their position and increasing their

wealth, a community of prosperous and progressive farmers with

little inducement to spend their time or their substance in

warfare.

South of the Thames a new situation had arisen. Commius,

for some years the servant of Julius Caesar, was among the

Gaulish chiefs who turned against him in the rebellion of Ver-

cingetorix, answering the last appeal of that gallant fighter and

staking all that they had on the attempt to raise the siege of

Alesia. When that attempt failed, Commius was in a desperate

position; but now that he had declared himself he was faith-

ful to the cause he had embraced, and became a dangerous

and effective enemy to Rome. During the final phase of the

Gallic war, he was at last beaten to his knees by Mark Antony,

but made his peace and gave hostages for his conduct on the

express condition that he should never again be compelled to

look upon the face of a Roman. This condition had reference

to a treacherous attempt on his life made at Labienus' com-

mand by Volusenus; but in the light of later events it was easy

to read into it another interpretation; for it is recorded that at

some date not specified Commius sailed with a fleet to Britain.

A writer of the following century tells a story of how, pursued

by the Romans, he escaped by hoisting his sails and seeming to

be already under way when in fact his ships were still aground

waiting for the tide; and when next we hear of him he is striking

coins inscribed with his own name in the woodlands of north-

eastern Hampshire.

Archaeology, filling in the details of this story, has shown that

some time about 50 B.C. there was a second migration of Belgae

into Britain. It seems, to judge from its pottery, based upon

56

FROM CAESAR TO CLAUDIUS

Normandy rather than upon the Marne valley; and it certainly reached Britain by way of the Hampshire coast. This second Belgic invasion was the movement led by Commius, who must have landed somewhere near Southampton, pushed his way up-country, and planted a city where now stand the ruined walls of Silchester.

In the course of time these western Belgae took possession of all the central part of southern England. From the Hampshire basin they pushed out in three directions. On the north their frontier is not accurately known; it probably ran along the escarpment of the Berkshire Downs and reached the Thames at Cholsey. Eastward they penetrated into western Surrey and, farther south, established themselves on the Sussex-Hampshire border. The South Downs had retained their primitive Hallstatt civilization almost untouched, and to the end they were uninfluenced by the eastern Belgic culture of Kent and the Thames basin; but now the old hill-fort of the Trundle was evacuated and a new city, Noviomagus, built in the plain on the site of Chichester, defended like many Belgic cities by cross-country dikes running at some distance from the town itself; Cissbury, too, on the hill above Worthing, was evacuated, and only eastern Sussex beyond the Adur remained independent.

On the west the realm of Commius absorbed by degrees all the chalk uplands of Wessex, and along their western edge built or rebuilt a formidable series of fortresses marking its frontier. They ran in a curved line roughly from Swindon to Bourne-mouth, and comprised over a dozen strongholds, crowning the crest of Marlborough Downs from Ashdown to Devizes and the escarpment of Salisbury Plain from Upavon to Warminster, and then, after certain intermediate forts near Wilton and Shaftes-bury, ending with a number in the Stour valley, of which the chief are the imposing twin works of Hod Hill and Hambledon near Blandford; and, crowning the whole, Maiden Castle.

As we shall see, this large tract of country was not always ruled by a single king. The immediate followers of Commius, drawn from the Gaulish Atrebates, were settled in and round Silchester; another tribal name, that of the Regni, maintained itself in west Sussex. But by the time of the Claudian conquest the civilization of the whole region was fairly uniform. It was a hybrid culture in which old and new elements were mixed. The old upland agriculture went on, for the most part in peace-

ful, unfortified villages; here and there, especially round Silchester, there was some clearance of forest and exploitation of heavier soils; and except along the western frontier, the practice of building or living in hill-forts appears to have died out as the country became more and more peaceful under the sway of a strong central government.

There were now two distinct regions of Belgic settlement in Britain, one in Kent and Hertfordshire, the other in and about Hampshire. The history of Britain between Caesar and Claudius is to a great extent the history of these two regions, and in attempting to reconstruct it we must begin with the history of the kings who ruled them.

The fortunes of the two royal houses can be to some extent inferred from their coinage. Their founders, Cassivellaunus and Commius, both struck coins. Those of Cassivellaunus, uninscribed but identified as his with reasonable certainty, are found mainly in the Catuvellaunian territory, Hertfordshire and the lands adjoining it north-westward; a similar type is found in west Surrey and west Sussex, and is difficult to explain. Contemporary with these or on average a little later are the coins of Commius, assigned to him with certainty because at this point coins begin to be inscribed. These are mostly found near Silchester, with outliers on the Sussex coast. It would seem that in the lifetime of Commius the northern and eastern frontiers of his realm ran down the middle Thames, diagonally across Surrey and through the Weald. The Catuvellaunian house ruled north of the Thames; so far as the coins of this period go, Kent seems to have lain in neither kingdom.

Cassivellaunus was succeeded by a ruler whose name appears on the coins as Andoco, doubtless an abbreviation. He, after a short reign, was followed by Tasciovanus, who thus appears as probably Cassivellaunus’ grandson. His coins are widespread and numerous; they indicate a long and prosperous reign, and, occurring as they do almost exclusively north of the Thames, chiefly in Hertfordshire, Bedfordshire, Buckinghamshire, and eastern Oxfordshire, but with scattered specimens over a wider area, they show the Catuvellauni becoming a wealthy and commercially active state about the end of the last century before Christ.

In quantity, the coins of the Commian house cut by comparison a poor figure; but they yield fairly definite historical

58

FROM CAESAR TO CLAUDIUS

inferences. It seems from the evidence of this coinage that Commius left three sons. Eppillus, at some time late in the century, carved out for himself a short-lived kingdom in north-eastern Kent. Tincommius at the same time ruled in Sussex and Hampshire, no doubt from the new city of the Regni at Novio-magus. Verica succeeded to his father's more immediate possessions round Silchester, and was the most successful of the three; for Eppillus was soon overthrown by a certain Dubnovellaunus, of whose origin we know nothing but whom we may conjecture to have been the last independent king of the Trinovantes, possibly even the son of Caesar's Mandubracus; and Tincom-mius, not much later, was evicted from his kingdom and sent into exile by Verica himself. Thus about the end of the century we appear to have three kings striking coins: Tasciovanus in Hertfordshire, the richest and most powerful of the three; Verica at Silchester, ruling in Berkshire, Hampshire, Surrey, and Sussex; and Dubnovellaunus at Colchester, commanding both sides of the Thames estuary.

Soon after the turn of the century the first and third of these kingdoms were united into one by the activity of Cunobelinus, son of Tasciovanus, who conquered the Trinovantes and ex-pelled Dubnovellaunus. During his long and extremely prosperous reign, which began about A.D. 10 and lasted until between A.D. 40 and 43, his coins appear not only over the whole area previously covered by those of Tasciovanus, but also over Essex, Kent, and eastern Surrey. With him the capital of the dynasty moves to Colchester, where his only mint was situated; his father had three mints, one at Verulamium and others at unidentified places whose names began with the letters Ricon . . . and Sego . . .

During the earlier part of Cunobelinus' reign Verica must still have been ruling over the now reunited kingdom of Com-mius. Throughout that kingdom coins of Cunobelinus are absent. But there are a few, scattered widely over its area, struck by Epatictus, another son of Tasciovanus; and it is not unlikely that this Epatictus was sent into the west to conquer Verica's realm and hold it as an ally of his brother Cunobelinus.

By this time the first half of the first century A.D. was well advanced, and three other British kingdoms were minting their own currency. One was the kingdom of the Iceni, northward neighbours of Cunobelinus. Their coins are derived from those

of Tasciovanus and are a good deal influenced by the mint of

Cunobelinus; many of them bear the name of Addedomarus,

who must have been king of the Iceni soon after the beginning

of our era,1 and evidently reigned for a considerable time.

Another was the kingdom of the Dobuni, whose coins spread

from a centre near Cirencester and Malmesbury over the Cots-

wolds, spreading eastward down the upper tributaries of the

Thames as far as Oxford, and westward across the Severn valley

into the Welsh foot-hills. Southwards they come in a wedge

through Somerset and Wiltshire into Dorset, with outliers in

Devon and Cornwall. Their eastern limit is marked on the

north by the western frontier of the Catuvellaunian kingdom,

and on the south by the western edge of the realm of Commius.

Six names of rulers appear on these Cotswold coins: one is

known to us in full as Antedrigus, the others in abbreviated

forms as Boduoc, Catti, Comux, Eisu, and Vocorio-ad. In

origin, this coinage is based on imitation of western Belgic issues,

especially those of Tincommius. It thus hardly begins to exist

before the beginning of the first century A.D.

The third kingdom outside the Belgic area to strike its own

coins was that of the Brigantes. These coins trace their pedigree

back to an carlier source than either the Iceni or Dobunian;

they seem to be an independent development from Continental

models not much later than those which produced the un-

inscribed coins of Cassivellaunus, and therefore testify to inde-

pendent connexions, whether by way of migration or trade,

between north-eastern Britain and the Continent. They occur

in Lincolnshire, Nottinghamshire, and Yorkshire.

In the generation immediately preceding the Roman con-

quest we thus find the more civilized part of Britain, on the

evidence of coins, divided into five realms: that of Cunobelinus,

extending from the Weald to the latitude of Cambridge and

from the Essex coast to the Cherwell valley; that of his brother

Epatictus, south of the upper Thames and the Weald and

extending westward to Salisbury Plain; that of the Iceni, con-

sisting of East Anglia and the Fens, and ruled by Addedomarus;

that of the Dobuni, stretching from the neighbourhood of

1 G. C. Brooke in an important article ('The Distribution of Gaulish and British

Coins in Britain', Antiquity, vii (1933), p. 268) has argued that the supposed Icenian

ruler Antedrigus never existed, and that the coins ascribed to him really bear, under

various forms, the abbreviated name of this same Addedomarus.

60

FROM CAESAR TO CLAUDIUS

Oxford and Banbury along the Cotswold ridge and including the low lands of Somerset and Dorset and a considerable sphere of influence in Wales; and that of the Brigantes, centred on the Humber and spreading loosely but widely over the north.

The only case in which the coin-evidence is unsatisfactory is that of the former Commian kingdom, here ascribed to Epaticcus. Actually his coins are very rare in it, though widely scattered from northern Oxfordshire to Selsey Bill; but coins of the Commian dynasty are rare, too, when compared with those struck by the house of Cassivellaunus; and most of the pieces found in this region are imitations of early Gaulish issues, progressively repeated and degraded, which in the western part of the area passed current to the complete exclusion of any inscribed coins whatever. This is one aspect of a more general fact to be discussed later in the present chapter, namely, the relatively primitive and backward character of this region during the century before the Claudian conquest.

Before leaving the subject of coins in pre-Claudian Britain it may be well to consider two other aspects of that subject: the general pattern of their distribution on the map, and the evidence for civilization given by their style.

When all the pre-Roman coins found in Britain are plotted on a single map, it is at once evident that, with trifling exceptions, they have been discovered to south and east of a line running from Weymouth to Bristol and thence along the oolite outcrop that forms successively the Cotswolds, the Northamptonshire uplands, and Lincoln Edge. Apart from a few scattered finds in the south-west, the only exceptions of any importance are the Dobunian coins in the Severn valley and Welsh foot-hills and the Brigantian ones in Yorkshire. The whole of the great midland triangle bounded by the oolite-ridge, the Pennines, and the Welsh mountains is utterly barren.

Within the area thus delimited they are much commoner than anywhere else in the district which ultimately became the kingdom of Cunobelinus. This comprises an area reaching from the Thames to the Ouse and Stour and from the sea to the Cherwell; south-eastward it includes all Kent except the Weald. Outside this district are other areas where coins are fairly dense, notably round Norwich and in western Suffolk; the Nene valley; the Cotswolds and the Oxford basin; north-eastern Hampshire and the adjoining part of Surrey; the Sussex coast; and the

neighbourhood of Poole and Cranborne Chase. On the positive side, it is worthy of remark that the Belgic penetration of denser forest-lands has gone far enough to appear on the map in the shape of thickly scattered coin-finds on the Chiltern ridge and round Silchester, to name the two most conspicuous cases, while on the negative side the most interesting fact is the scarcity of coins on Salisbury Plain. We shall have cause to notice, later on, the curious way in which Salisbury Plain resists new influences throughout the Roman period; it is interesting to observe that it has already taken up this attitude before the Claudian invasion.

The style of the ancient British coinage is chiefly interesting, for our present purpose, as an index of Romanization. This coinage began by imitating that of the northern Gaulish tribes, and this again originated in copying the gold staters of Philip II of Macedon which began to flood the Roman market and penetrate freely into north-western Europe towards the middle of the second century before Christ.1 The Philippic stater has on the obverse a head of Apollo with a laurel wreath binding its tightly curled hair, and on the reverse a two-horse chariot. The copies of this coin which became the model for the British moneyers were struck by the Atrebates and other tribes of Belgic Gaul. The general characteristic of all these models is that the Gaulish craftsmen, unaccustomed as they were both to the technique of die-sinking and to the whole conception of representing human or animal forms naturalistically, reduced both designs to what, from the point of view of naturalistic art, can only be called nonsense. On the obverse the laurel-wreath has become a band of pellets somewhat resembling an ear of wheat; on either side of this are tassels, hooks, or bosses representing the locks of hair, and strange crescent-like forms derived from the outline of the face. On the reverse the two horses have been reduced to one, and that is strangely distorted and runs some risk of falling in pieces; the chariot and other accessories are becoming mere dots and streaks, muscae volitantes, in the field.

1 G. C. Brooke in Antiquity, vii (1933), pp. 268–9, argues forcibly, in opposition to the generally received view, that it was such events as the triumph of Scipio Asiaticus in 188 B.C. and the battle of Pydna in 167, with their enormous spoils in gold staters, that brought these coins into the western market: and further, that it was the victory of Ahenobarbus over the Arverni in 121 that introduced them into central Gaul. The general chronology of the Gaulish coins is rendered more intelligible by these suggestions.

62

FROM CAESAR TO CLAUDIUS

When the Britons took this coinage as a model for their own

the Apollo-head had advanced so far in disintegration that it

was past recovery and was nothing but a tangle of disjointed

elements. The British die-sinkers, whether consciously or un-

consciously, set themselves the task of reducing this tangle to

order, modifying it very gradually in the direction of a decora-

tive pattern. The clement which had suffered least in the

disintegrative process was the laurel-wreath, with its tendency

to become something like an ear of wheat; the course usually

taken by the British moneyers was to duplicate this at right

angles, making a cruciform design, in whose centre they col-

lected crescentic motives derived ultimately from Apollo's pro-

file. In the time of Tasciovanus, under whom this development

reached perhaps its highest point, this central group of motives

framed a little panel on which the king's name could be in-

scribed.

At the same time, the degeneration of the horse is arrested.

The Britons, like Artemus Ward, found themselves unable any

longer to conceal from the public that these things were horses,

and, simultaneously with the attempt to organize the scattered

elements of the obverse into a single decorative design, there

is visible in their work an attempt to rehabilitate the horse on

the reverse, infuse it with new life drawn from actual observa-

tion, and produce what at its best is not only a recognizable but

a vigorous and almost naturalistic animal portrait. This was

the general character of the British coinage as developed under

Cassivellaunus and taken over by Tasciovanus. A somewhat

similar character belongs to the issues of Commius and to the

Dobunian coinage derived from them, though here an inde-

pendent connexion can be traced with Gaulish sources.

A new movement sets in towards the beginning of our era.

Already under Tasciovanus silver pieces, though not gold, begin

to imitate contemporary Roman work. This tendency rapidly

gathers weight. It reaches its highest pitch of intensity in the

reign of Cunobelinus, whose silver and copper is entirely domi-

nated by it, and even his gold strongly affected. About the same

time the same change is taking place in the realm of Commius.

Of his three sons, only Tincommius uses the traditional British

designs, and he changes over to the Romanizing fashion before

the end of his reign; Verica and Eppillus invariably strike in

the new style.

ARTS AND CRAFTS

63

This new style, then, dominates the coinage of all south-eastern Britain during the first half of the first century A.D., and towards the close of that period it even begins to affect that of the Dobuni. Broadly speaking, it may be described as a Romanizing style, and at times it may even have been practised in Britain by craftsmen imported from the Continent; but in spite of the profusion of types borrowed from Roman numismatic art, it retains a strongly Celtic colouring. Side by side with heads closely imitated from those of Augustus are others, whose savage and hirsute appearance contrasts curiously with the skill employed in their portrayal. Roman fashions are penetrating peacefully into British society; but that society itself, though welcoming these luxuries as an adornment of its life, is at heart barbarian still.

Much more interesting than the coinage, but at the same time much harder to interpret historically, owing to the difficulty of dating its products, is the artistic work of the Britons during the generations preceding the Claudian conquest. The products in question are, at their best, the work of highly accomplished artists, gifted with a sense of line that has seldom been surpassed in any school, and employed to decorate objects of luxury evidently belonging to the wealthy and warlike aristocratic class of the community: shields, helmets, scabbards, mirrors, personal ornaments, and horse-trappings. These metal goods have survived in small numbers where a vast quantity of more perishable work has evidently been lost; consequently they come to us without any context, and this is the chief reason why they are so hard to interpret. It is well, therefore, to turn first to the humble domestic craft of pottery, which may provide a background for these higher works of art.

Of the ruder pottery little need be said. All over south-eastern England it is based mainly on the Hallstatt tradition of the original Iron Age settlers; elsewhere the Bronze Age tradition still predominates, where pottery is found at all. Against this background there stand out three1 main types of better and more civilized wares.

First, there is the decorated hand-made ware which is best known to us at the Glastonbury lake-village. The interesting feature of this ware is its decoration. Derived from Gaulish art of the middle La Tène period, it delights in boldly drawn curved

1 Of the first and second something has already been said in Chapter II.

64

FROM CAESAR TO CLAUDIUS

lines, spirals, and circles, S-shaped curves, wavy, zig-zag bands,

and chains of lenticular or almond-shaped elements, all these,

at the same time, crossed and combined with decorative motives

drawn from the repertory of the Bronze Age. The predominant

shape for vessels is that of a squat, wide-mouthed jar or jar-like

bowl. Pottery of this kind is widely thought somewhat thinly

spread over the south-west, from Cornwall to Hengistbury Head

and even penetrating into Sussex; inland it spreads over Dorset

and Somerset and works its way along the Cotswold ridge as

far as Hunsbury in Northamptonshire. It thus attests a cultural

unity belonging to the same region whose central part is charac-

terized by the Dobunian or Cotswold coinage. This culture

goes back for its origins, as has been said in an earlier chapter,

to the fourth century before Christ; but we are now considering

its fortunes after the invasion of Julius Caesar, and here the fact

of chief importance is the destruction of the Glastonbury village,

apparently by raiding Belgae from the realm of Commius, at

some date between that time and the Claudian conquest.

During this period, then, the Dobunian civilization was being

attacked by the Belgic; but the evidence of its coins proves that

on the whole it maintained its independence.

Secondly, there is the wheel-made pottery of the original

Belgic invaders, also derived from Gaulish La Tène models, but

differing from the south-western wares in its utter lack of decora-

tion. It is not altogether inartistic, but it produces its somewhat

dry and hard effect by its elegant and decisive profiles, con-

trasting strongly with the squat shapes of the south-west. This,

the Aylesford pottery mentioned in a previous chapter, extends

between the time of Caesar and that of Claudius all over the

region that has already been defined in the present chapter as

the kingdom of Cunobelinus.

Thirdly, there is a type of pottery called by archaeologists

bead-rim ware, which is found throughout the realm attributed

to Commius and his family. This is regarded as the pottery of

the second Belgic invasion, to some extent certainly influenced

by fashions already existing in that area, but in the main de-

rived from wares used in Normandy during the last century

before Christ. It is sometimes hand made, sometimes wheel

made; it is undecorated, except when it is rudely scratched with

a comb; and it shows no evidence of artistic talent in its makers.

The only region, therefore, in which an art akin to that of the

finer metal-work has penetrated down to the level of domestic

pottery is the south-west with its prolongation into Northamp-

tonshire. Now, the La Tène art of Britain may perhaps be seen

at its best, both in quality and in quantity, in the engraved

bronze mirrors which have always been recognized as a pecu-

liarly British class of object. Over a score of them are known;

some are too fragmentary to give an idea of their original

appearance, but several are almost or quite perfect. Two

of them stand out as among the very best works of Celtic

art: they were both found in the region of which Glastonbury

pottery is characteristic, one at Birdlip in Gloucestershire, the

other at Desborough in Northamptonshire. The rest, so far as

their condition enables us to reconstruct them, are debased and

degenerate copies of the beautiful designs happily preserved to

perfection in these two. If we seek for parallels to the decoration

on the Birdlip and Desborough mirrors we shall find them

primarily in the Middle La Tène ornament found at sites like

Meare in Somerset, Hunsbury in Northamptonshire, and the

country of the Brigantes, and secondarily, simplified into the

relative crudity of peasant-art, on pottery in the Glastonbury

style. Thus the evidence concerning the artistic culture of the

entire belt of country that begins in Somerset, and travels along

the oolite-ridge into Lincolnshire and Yorkshire, falls into an

intelligible whole. It was into this belt of country that the best

elements of Middle La Tène art were transplanted; it was here

that they took root and, especially perhaps in the southern and

central part, blossomed into the school of British art that gave

us these mirrors.

It remains to ask when this blossoming took place.¹ The

difficulty of dating such works of art has already been confessed.

The Birdlip mirror is usually dated to the middle or second half

of the first century A.D., that is, shortly after the Roman con-

quest, because in the same grave with it was found a unique

and beautiful brooch which has been thought to be derived

from types of brooch worn in Germany about A.D. 40–60. In

fact, however, this brooch-type seems to have been decadent

¹ The reader ought to be warned that in this paragraph a view is maintained

which many archaeologists would reject. They would place the Birdlip mirror after

the conquest, on account of the associated brooch. I have long regarded the argu-

ment as weak and the conclusion as impossible, and am glad to have my doubts

shared and confirmed by E. T. Leeds, Celtic Ornament in Britain (1933), pp. 28–37, to

whose statement of his case I am much indebted.

66

FROM CAESAR TO CLAUDIUS

during those years, and soon afterwards it passed out of use; and

there is no evidence to tell us how long before this it might have

influenced a British artist. We are on firmer ground in the fact

that a debased and presumably late mirror was found in a grave

at Colchester with pottery of the late first century B.C. The

natural inference is that the school of art which gave us the

Birdlip and Desborough mirrors culminated about the time of

the Julian invasion.

Thus, in the early part of the century which we are now at-

tempting to describe, the regions of Belgic settlement may have

been in some ways, as Caesar thought, the most civilized parts

of Britain; but just beyond their limits lay a region less pros-

perous perhaps in its agricultural basis, but far richer in its

artistic culture, and supporting a wealthy and powerful aristo-

cracy with a purely Celtic tradition of life and art. The north-

westward movement of the Catuvellauni after Caesar's invasion

brought them into contact with this Celtic region; and as the

wealth and power of the Catuvellauni increased they began to

develop an art of their own, based chiefly on borrowings from

it. Inferior copies of its mirrors began to be used in the Belgic

territory, and as time went on a new school of metal-work grew

up there, which made important technical advances in the craft

of enamelling.

But the taste of the Belgae for Celtic art was never very strong.

At the time of their greatest prosperity under Cunobelinus their

chief men were deserting Celtic fashions for Roman. Of this we

have much evidence gleaned from various sites of which the

richest, and also the most decisive, is the magnificent burial-

mound of Lexden near Colchester. The ritual here is British;

may not be the grave of 'the radiant Cymbeline' himself) was

apparently buried with his wår-chariot; the pottery is Belgic,

but the metal-work, of which great quantities were found in the

tumulus, is wholly Roman either in origin or in taste, and there

is a complete absence of any trace of Celtic art. Elsewhere in the

realm and age of Cunobelinus the same tendency to desert Celtic

fashions for Roman, though never so overwhelming, is plainly

visible: throughout the first half of the first century of our era

the evidence of burials seems to show a progressively romanized

court and aristocracy, soon parting with the Celtic fashions they

had but lately and partially acquired.

THE BRIGANTES

67

In the realm of Commius there was indeed a less intense romanization, but there was also far less celticizing. Its capital,

Silchester, was already using small quantities of earthenware from Italy before the Claudian conquest; but in many years'

excavation practically nothing has been found bearing any resemblance to La Tène art. Like their more powerful neigh-

bours and kinsmen to east and north-east, these western Belgae of the second invasion were an inartistic race, but, unlike them,

they never advanced so far in wealth and power as to attract the arts of luxury; and they remained to the end a people of

simple farmers and, on occasion, stout warriors.

Meanwhile the north, free from Belgic influence, was develop-

ing on lines of its own. There are here no villages with a true Iron Age culture, whether of Hallstatt or La Tène type, no hill-

forts like those of the south, and no towns to compare with Verulamium or Cunobelinus's capital at Colchester. Against

all these we can only set a single site,1 the vast oppidum at Stan-

wick near Darlington: a site where a tract of plain country, a mile by a mile and a half in extent, is enclosed by a great ram-

part and ditch of regular Iron Age pattern, fit capital, one might conjecture, for a warrior-state embracing in its dominion all

northern England. Whether or no that is the true explanation of the Stanwick earthworks, the descendants of the Middle La

Tène invaders had by now grown into a Celtic warrior-caste ruling over a populace of primitive culture hardly differing from

that of a poor and ill-furnished Late Bronze Age; and thus had come into existence the state of the Brigantes or hillmen. A

similar process appears to have taken place in south-eastern Wales, Monmouthshire, and Herefordshire among the Silures,

where, however, the influence of Dobunian culture led to a considerable development of hill-forts.

The west, where pre-Roman Celtic art in Britain reached its zenith, seems hardly to have retained its supremacy in this

respect until the time of Claudius. Its northern extension on

1 Yorkshire contains a number of small hill-forts, for which see Elgee, Archaeology

of Yorkshire (1933), pp. 117-19; they are quite unlike the great forts of the south, and may have been constructed in the course of the Roman invasion. It is other-

wise with Stanwick, where the rampart and ditch (100 feet across, and still some 30 in vertical height) must represent a huge permanent fortress, and cannot (pace Elgee, op. cit., pp. 233-4) be a medieval deer-park. The La Tène hoard in the

British Museum confirms this view, whether found inside or just outside the rampart. At Almondbury, near Huddersfield, is another fortress of Early Iron

Age type which ought perhaps to be mentioned together with Stanwick.

68

FROM CAESAR TO CLAUDIUS

the Northamptonshire uplands was already being encroached upon by the Catuvellauni before the turn of the century. Its southern extension was threatened by the fortresses of the Belgic frontier, and almost certainly harried by Belgic raids, not very much later. Even in its centre we find, here and there, evidences of Belgic penetration into the very fortresses of the Cotswolds.

Its coinage, which falls, as we have seen, almost wholly, if not altogether, after the beginning of our era, shows no evidence of artistic vitality. And when we are able to recapture the spirit of this region once more in the artistic work done there soon after the Roman conquest, we find it, with one notable exception to be considered hereafter—the sculptures of the Bath temple—strikingly inferior to what is being done elsewhere. If it is true that the Birdlip mirror was made not later than the last century before Christ, we have here a consistent picture of the art-loving aristocracy of the west as decaying, whether from internal or external causes, during the first half-century of our era. And this was the time when the leadership in the field of art passed to the north, and was taken up by the Brigantes.1

The economic life of the country, especially in the Belgic areas, shows a certain development during this period. Of agriculture little need be said beyond what has been said already. We have seen that in addition to the village-communities with their little fields there were already isolated farms presumably ploughing large tracts of unenclosed land, and that the latter system may possibly have been characteristic of the Belgae with their heavy sod-turning plough. We have seen that large quantities of wheat were grown, and it may be added that barley also was grown and used for brewing beer, the national drink of the Celts in every land. It need hardly be said that Caesar was misinformed when he was told that, except near the Straits of Dover, the Britons knew nothing of agriculture but lived on meat and milk and dressed in skins. Even the remoter and less civilized tribes, long before his lifetime, were cultivating the soil and wearing woollen and linen clothes.

On another point, in spite of common opinion to the contrary, Caesar appears to have had better information. He says that the Britons in his time had to import their copper.2 As Britain

1 I must once more express my debt here to Leeds, whose demonstration (Celtic Ornament, cit., pp. 45–8) of decadence in western art before the conquest, as exemplified by the latest pieces in the Polden Hill hoard, is most convincing.

2 Aere utuntur importato (B.G. v. 12, 4). Aes may mean (1) copper, (2) bronze, the

is somewhat rich in copper, and has been a source of bronze

from very early days, it has generally been thought that Caesar

was wrong here. But in point of fact there is no evidence that

any of the copper ores in Britain were mined before the Romans

exploited them, and it is by no means impossible that the impor-

tant Bronze Age foundries of this country, though deriving their

tin from native sources, drew their copper from abroad, and that

this state of things lasted until the Roman conquest.

Iron was being worked in large and increasing quantities,

mostly, no doubt, still in the Weald, but also perhaps in the

Forest of Dean, almost certainly in the Mendips, and probably

in many other places also. The large hoards of currency-bars

found in the south-west and the frequency with which iron

objects of considerable size, like fire-dogs, were used during this

period mark it as an age when iron was easy to come by and

when the working of it was an art very familiar to blacksmiths

in all parts of the country.

The Romans in the time of Claudius expected Britain to

provide gold in payable quantities; and certainly it was being

exported thence in the reign of Augustus. The Wicklow deposits,

it is thought, had been worked out some centuries earlier, and

therefore other sources must have been known. Possibly these

were in South Wales; possibly in Cornwall; possibly in Scotland;

the first alternative is the most probable, but the other districts

may have produced gold also.

Of silver we hear nothing in Caesar's notes on Britain, and

Cicero (as already quoted) writes to a friend that there was

none. In the middle of the last century before Christ the great

wealth of Britain in argentiferous lead was as yet unknown; it

seems that none of the deposits had been worked except those

of the Mendips, and they not on a large enough scale to attract

attention. The lake-village of Glastonbury proves, however,

that the Mendip lead was being regularly mined in Caesar's

time and before; it has yielded several leaden objects, mostly

net-sinkers, and others have been found in the earlier village,

not far away, at Meare; while at the mining village of Charter-

house British coins have been found. The use of lead for such

alloy of copper and tin, (3) coinage. The context excludes this third meaning. It

has also been suggested that the word here means 'bronze-ware'—saucepans and

the like. It is certainly true that after Caesar's time much bronze-ware was im-

ported into Britain; but that is not the natural meaning of the passage, which must

refer to the metal, either pure or alloyed.

70

FROM CAESAR TO CLAUDIUS

everyday purposes as weighting fishing-tackle proves that by

now it was cheap and common. The lead from Glastonbury

and Meare has not been desilverized; but this process must have

been introduced into Britain at an early date, if it is true, as

Strabo says, that by the time of Augustus silver was being

exported from Britain on a considerable scale.

The early British tin-trade is so controversial a subject that

some hardiness is needed in order even to mention it. Fortu-

nately, we are here concerned only with the half-centuries

immediately before and after the birth of Christ. We need not,

therefore, discuss the difficult geographical problems—Cassite-

rides, Ictis—which surround the prehistoric tin-trade. In

Caesar’s time Cornwall was still the source of most of the tin

used in western Europe and the Mediterranean, but the secret

of its origin was so well kept that Caesar himself only knew that

it came from the ‘inland regions’, by which he evidently meant

some region not near the Straits of Dover. Actually the carrying

trade in his own times was in the hands of the Veneti, and what

happened to it after the destruction of their shipping we can

only guess. But during the generation that followed his British

campaign the Augustan settlement of Spain opened up the

Spanish tin-mines to the Roman world, and there is little doubt

that this event was a severe blow to Cornwall. Nevertheless, the

industry continued; and soon after the conquest, as we shall see,

Roman prospectors in their systematic search for minerals found

their way to Cornwall and settled there for a time.

Whitby jet and Kimmeridge shale, out of which ornaments

were turned on the lathe and carved, were already staple pro-

ducts of Britain and were widely used throughout this period.

Pearls, too, were an article of commerce; Caesar himself is

recorded to have made an offering to Venus Genetrix of a

breastplate set with them.

There was, in fact, during this century, a large and growing

trade between Britain and the Continent. Caesar’s campaign

against the Veneti must have crippled the Atlantic coastal

traffic; but even in his own time, he tells us, that which crossed

the Straits of Dover was more important, and this increased

steadily in bulk and value. During the reign of Augustus

Strabo enumerates the exports from Britain as comprising

wheat, cattle, gold, silver, iron, hides, slaves, and hunting dogs;

under imports he mentions bracelets, necklaces, amber, glass-

RELATIONS WITH THE ROMAN EMPIRE

71

ware, and ‘suchlike trifles’; and explains that the Romans have

abandoned the idea of annexing Britain because, if they did so,

they would perforce lose the large revenue collected in the form

of duties on this trade. To this argument we shall have to

return when considering Roman policy towards Britain in this

period. For the present it may be observed that, so far as can

be judged from the evidence of archaeology, the argument by

no means exaggerates the extent or value of this cross-channel

trade. Indeed, Strabo rather under-estimates it. Amphorae in

which wine must have been brought to Britain from the Roman

world are found in such large numbers that the wine-trade must

have been very considerable. Arretine pottery was coming into

general use among the wealthier classes all over the Belgic

regions, especially at Colchester itself. The Campanian metal-

work, both of bronze and silver, found in tombs at Lexden and

elsewhere in the realm of Cunobelinushas already been men-

tioned. Strabo is clearly right in implying that what Britain

imported from the Roman world was articles of luxury; but of

these, as the romanizing movement of the last two generations

before the conquest went forward, the variety, the bulk, and

the value steadily increased. It is not surprising to learn that

during the reigns of Augustus and Tiberius a trading-settlement

grew up on the left bank of the Thames, on the site of London,

where goods from Gaul and Italy were handled in considerable

quantities. London was already entering upon its career as the

leading commercial port of the British Isles.

It remains to consider the relations during these hundred

years between Britain and Rome.

If Caesar intended to conquer Britain, and if his intention

was at Rome an open secret, it is easy to understand an other-

wise perplexing fact, namely, that Augustus is repeatedly

credited by good authorities with the same intention. Horace

states it as all but a fait accompli:

Caelo tonantem credidimus Iovem

Regnare; praesens divus habebitur

Augustus, adiectis Britannis

Imperio gravibusque Persis.

Dio alleges that in 34 Augustus actually set out to invade

Britain ‘in order to outdo the feats of his adoptive father’, but

abandoned the project because of a revolt in Dalmatia. In 27,

72

FROM CAESAR TO CLAUDIUS

Dio tells us, he again started for Britain, but remained in Gaul partly because Gaulish affairs were unsettled and partly because he expected envoys from Britain, who never came. A third intended expedition is ascribed by Dio to the year 26; its cause, the refusal of the Britons to come to terms; the cause of its abandonment, the revolt of the Salassi and the outbreak of a serious war in Spain.

These stories as Dio tells them are circumstantial enough; confirmed by the testimony of Horace, we cannot dismiss them as groundless fictions. Their meaning, especially in view of the words about outdoing Caesar, is plain: they refer to a project of conquering Britain and annexing it, as Horace says, to the Empire. They prove beyond a doubt that, up to the year 26 B.C., if no longer, the intention of conquering and annexing Britain was generally attributed to Augustus, and that he never disavowed it. Whether he entertained it is quite another question.

Augustus had none of Caesar's adventurousness. He was the most circumspect of men. It was most unlikely that he would seriously think of committing himself to the conquest of Britain while so many urgent problems awaited solution in Gaul, in Spain, in the Illyrian provinces, and in the East. The most probable explanation of the stories that have come down to us is that Augustus, knowing, as every one knew, that his adoptive father had meant to conquer Britain and had been prevented from pursuing his plans by trouble in Gaul, allowed it to be assumed that, with the rest of his inheritance, he had taken over this project too; and it was no doubt gratifying to his rather malicious sense of humour to find that whenever people expected him to go ahead with this particular piece of work, some variant of the old excuse—trouble in Gaul—was always at hand to explain why nothing was done.

But these excuses could not be repeated indefinitely. A time came when every one must realize that Augustus had no intention of annexing Britain; and, always on the assumption that this was known to be part of Caesar's plans, a reasoned explanation was needed for the supposed change of policy. The explanation which must surely have been the official one has been recorded by Strabo in two passages. In the first he argues that

1 Horace, Odes iii. v. 1–4 (cf. i. xxxv. 29). Vergil has references to Britain which, though less explicit, confirm those of Horace. Dio, xlix. 38. 2; liii, 22. 5 and 25. 2.

2 Strabo, ii. 115–16; iv. 200.

Britain is so difficult of access that it can do Rome neither harm nor good: it is not strong enough to invade her territory, and the duties on the trade between it and Gaul bring in more than the total estimated yield of the tribute minus the cost of maintaining a garrison. In the second, the financial argument is repeated with further details, but for the argument about difficulty of access another is substituted, namely, that some of the British rulers have made their peace with Augustus, dedicated offerings in the Capitol, and ‘made the whole island almost a Roman country’.

There is in all these arguments a certain strain of disingenuousness. The financial plea looks like a repetition of what Caesar had to say concerning the poverty of Britain; but by the time of Augustus it was known, and these very passages of Strabo prove it, either that Caesar was mistaken or that times had changed. A country which could afford so large a trade with the Empire could pay its way in taxes as a province of the Empire. The plea that Britain was to all intents and purposes inaccessible is equally inconsistent with the recorded fact of constant trade relations.

As to the argument from the alleged friendliness of British princes, it can be checked by reference to Augustus’s own autobiography. In his Res Gestae Augustus records with pride the names of various countries outside the Empire whose rulers had sought his friendship; and the name of Britain is not to be found there. All he can say is that he had been visited by two suppliant, and therefore doubtless exiled, princes from that country, a certain Dubnovellaunus and some one whose name begins with the letters Tim . . . or Tin . . . . These are clearly no others than the Dubnovellaunus and Tincommius of the British coins. When Augustus wrote that autobiography, Cunobelinus had been on the throne for some ten years; and Augustus can only be tactfully concealing the fact that, during all this time, he had never sent him so much as a complimentary message.

The British rulers whom Strabo mentions are obviously the suppliant princes of the Res Gestae; and their relations with Augustus are thus evidence not of British friendship but of British hostility, a hostility quite compatible with advancing romanization in manners.

But Cunobelinus, though he never sought the friendship of Rome, never did anything to excite her enmity. He must have been well aware that, if war should break out, a single campaign

74

FROM CAESAR TO CLAUDIUS

would suffice to destroy him. His attitude towards Augustus

was therefore one merely of reserve; and when, during the

northern campaigns of Germanicus, certain Roman soldiers

were shipwrecked on the coast of Britain, it was quite in accor-

dance with his policy that the princes in whose country they

landed should send them safely back to the Continent.

So far as we can divine the policy of Augustus, British con-

quest had no part in it. He planned and carried out a thorough

pacification of Gaul, suppressing revolts and paving by all possible

means the way to peace; when this was done, he embarked on

his one great military failure, the attempted conquest of Ger-

many. By the time this attempt was renounced, after the slaugh-

ter of Varus's legions, we hear no more of Britain; the arguments

preserved by Strabo had done their work. Yet Augustus un-

doubtedly knew what the consolidation of Cunobelinus's king-

dom meant to Rome. He can hardly have been blind to the

fact that across the Channel a strong state was growing up, pro-

gressive in its civilization and by no means friendly to Rome in

its spirit: a state whence no overtures of friendship were forth-

coming, except from its exiles. `He is not likely to have been

deceived by the arguments which he used in explaining to others

why the conquest of Britain was not being pushed forward; and

in that case he must have known that, sooner or later, it was

inevitable. Either Rome would find her hands free to settle an

account long outstanding, or something would happen to pre-

cipitate a conflict.

Meanwhile he bequeathed to his successor the maxim that

the frontiers of the Empire were best left unaltered, and in

particular gave him a hint, which Tiberius treated (according

to a phrase of Tacitus) as a command, against reopening the

question of a British war. The new reign was a time of peace

and consolidation, and saw no change in the relations between

Britain and Rome. But events on either side were converging

towards a crisis. The very success of Tiberius's policy paved

the way for its own abandonment. The Rhine and Danube

frontiers were firmly based; the east was quiet; Spain, pacified

at last, was patently over-garrisoned; there were troops to spare

for a forward move wherever one was needed. And in Britain

Cunobelinus was growing old and his sons were quarrelling. In

the light of later developments we can discern a pro-Roman

and an anti-Roman party at his court, the latter led by two

THE ABORTIVE INVASION OF GAIUS

75

of his sons, Togodumnus and Caratacus, the former led by

another son, Adminius.

By the time Gaius came to the throne in 37, it was a perfectly

serious question of policy whether the time had not come for

undertaking the long-deferred conquest. Unhappily, Gaius was

a conceited and cruel tyrant upon whom his contemporaries

have revenged themselves by depicting him to posterity as a

lunatic; and it is impossible to discover the truth even concern-

ing his actions, not to speak of his intentions, beneath the mass

of ridicule that ancient writers have heaped upon them. In all

probability Gaius had resolved to invade both Germany and

Britain within a year of his accession; but the events which

brought him to the Rhine in 39 arose not out of a planned

campaign but out of a conspiracy against his life, which he

hastened to Germany to suppress. The strange stories told of

his doings there during that autumn have been explained with

some plausibility1 as a garbled version of manœuvres for the

training of the troops in view of an intended invasion of Ger-

many next year; and the fact that this invasion never took place

is accounted for by the same writer as due to the arrival of

Adminius, exiled by his father, bringing with him a handful

of followers, and coming as a suppliant to Gaius as he lay

encamped at Mainz, an event which, it is suggested, diverted

his campaign of A.D. 40 from Germany to Britain.

But in attempting to plumb the motives of Gaius we are

groping in the dark. All we know is that he did plan an invasion

of Britain for the year 40; that in preparation for it he had a

lighthouse built at Boulogne; that he assembled his expedi-

tionary force there, made all ready for departure, and then,

at the very moment of embarkation, cancelled his orders and

countermanded the whole campaign. The plan had been hastily

formed by an ambitious and inexperienced ruler; whatever may

have been the motive for its abandonment at that critical

moment, the fact that it was never taken up again is probably

due less to his own instability of purpose than to the atmosphere

of opposition and conspiracy with which he was surrounded at

Rome, an atmosphere whose ill effect on his character is ad-

mitted even by his kindliest biographers. To the historian of

Britain, however, the incident is crucial. It marks the end of

the long respite from Roman invasion.

1 J. P. V. D. Balsdon, The Emperor Gaius (1934), pp. 78–82.

BOOK II

THE AGE OF CONQUEST

V

THE CLAUDIAN INVASION

To the question why Claudius resolved to conquer Britain, ancient writers give no clear answer. Dio, to whom we are indebted for the only narrative we possess of the invasion, says that a certain Bericus, exiled from Britain, persuaded Claudius to attack it. But it is hardly probable that a Roman emperor undertook such an enterprise out of kindness to an exiled prince; nor does the name of Bericus reappear at any later stage in the narrative. Suetonius, the official scandal-monger of early Imperial history, alleges that the emperor was moved by the desire to celebrate a triumph; but in the same breath he observes that Britain was in a disturbed condition (tumultuantem is the word), which provides a motive of a quite different kind for Claudius's act: a motive based not on personal grounds but on grounds of imperial policy. Tacitus, by far the greatest historian of them all, hints that the mineral wealth of Britain had something to do with the decision; but in the brief summary of events, prefaced to his biography of Agricola, he conveys to his reader the impression that what demands explanation is not the fact that Claudius conquered Britain, but the fact that no one else had done it. For, says he, Julius Caesar had fought a successful campaign there, but in effect had only pointed Britain out, not handed it over, to his successors: pointed it out, we are to understand, not as an object of disinterested speculation or scientific study, but as a field for future conquests. Civil wars and political problems then, Tacitus continues, occupied for a while the attention of emperors; even in times of peace Augustus thought it politic, and Tiberius treated it as a command laid upon him by Augustus, to forget that Britain existed. Gaius's invasion was rendered futile by his own inconstancy and his record of ill success in Germany; it was left for Claudius to take up the task again and carry it through.

The question which Tacitus is here assuming, as in the mind of

1 Tacitus, Agricola, xiii.

REASONS FOR THE INVASION

77

any reader, is not 'Why did Claudius invade Britain?' but 'Why

was so obvious and natural a step not taken earlier?'

There is no doubt that the perspective in which Tacitus here

places the Claudian invasion is the right one. The project of

conquering Britain had been brought by Caesar into the agenda

of Roman policy. Thereafter, the question at any given moment

was not whether it should be done, but whether it should be

done now, and if not, why not. Augustus, as we have seen, had

more urgent things to do, and gave it out from time to time that

Britain was too poor, too distant, or too friendly to make

annexation expedient or necessary. By the time of Gaius none

of these reasons held good. When Claudius reviewed the situa-

tion, the facts which must have weighed with him may be

summed up as follows.

First, there were troops to spare. The Rhine, the Danube,

and Spain were amply guarded, perhaps even excessively. The

military situation is rendered a little ambiguous by doubt

whether it was Gaius or Claudius who enrolled the two new

legions, XV Primigenia and XXII Primigenia, which, with

IV Macedonica from Spain, took the place of the three Rhenish

legions sent by Claudius to Britain: II Augusta from Strasburg,

XIV Gemina from Mainz, and XX Valeria Victrix from

Cologne. If the new legions were recruited by Claudius,1 the

Rhine garrison was maintained at its former strength and its

composition changed; if by Gaius, it was done in preparation

for his own schemes of conquest, and the fact of its being done

meant that unless Claudius carried out those schemes the Rhine

would be heavily over-garrisoned. The fourth legion for Britain,

IX Hispana, was drawn from Pannonia.

Secondly, the fiasco of Gaius's abortive invasion laid a respon-

sibility on his successor. Whether Gaius was right or wrong in

thinking the time favourable for conquering Britain, the fact

that he had taken the first step towards conquering it gave

Claudius an additional reason for taking up the plan again and

carrying it out decisively, as a demonstration of firmness both

to the Britons and to his own legions.

Thirdly, the financial arguments of Augustus were long out

1 H. M. D. Parker, The Roman Legions (1928), pp. 93-8. Contra, E. Ritterling

in Pauly-Wissowa, s.v. 'Legio', 1244-9; Syme in Camb. Anc. Hist. x. 788-9. It may

be convenient for some readers here to be referred on to p. 138, where they will

find details about the strength, &c., of legions and auxiliary units.

78 THE CLAUDIAN INVASION

of date. Britain was known to be rich in raw materials, metals,

wheat, cattle, slaves; there was good reason to think that, as a

province, it would pay its way. Annexation would further

stimulate trade, and commercial interests of every kind must

have favoured it.

And lastly, the political situation was precarious. Even in the

time of Gaius, the expulsion of Adminius probably indicates

that the cautious and moderate anti-Roman policy of Cuno-

belinus was being replaced, as he grew old, by a more violent

version of itself, championed by Togodumnus and Caratacus.

By the year 43 Cunobelinus was dead and the anti-Roman

movement had come out into the open. The exile of Bericus

marks a further stage in its development: it appears that the

British princes demanded of Rome that he and perhaps others

should be handed back to them, and when this was refused they

took up a threatening attitude. Because Gaius's expedition had

never sailed they thought themselves secure from invasion, and

fancied that Rome could be bullied. Threats conveyed through

diplomatic channels might easily be followed by hostile action:

perhaps a descent on the coast of Gaul, more probably a mas-

sacre of the Roman traders who must by now have been settled

in Britain in considerable numbers. The case for invasion was

complete.

It has been said that Claudius formed the nucleus of his

expeditionary force by detaching three legions from the Rhine

garrison and one from Pannonia; it was from Pannonia that he

drew its commanding officer, Aulus Plautius. With the comple-

ment of auxiliary regiments, foot and horse, usual at that time,

the total must have numbered about 40,000. As compared with

Caesar's army in 54, there was one legion less but more auxi-

liaries; the chief effect of this would be that the Claudian force

was stronger in cavalry and therefore better adapted to the

tactical conditions of British warfare.

This is not the only respect in which the Claudian invasion

bears traces of planning by men who had carefully and profit-

ably studied Caesar's account of his own. The same point of

departure was chosen, at Boulogne; the same method of crossing

at night; the same landfall, with one significant modification,

in east Kent; and the same strategy for the opening stage of

the campaign.

Another resemblance, though an undesigned one, was that, as

THE EXPEDITION SAILS

79

Caesar found some of his Gaulish companions reluctant to cross

the sea, so Claudius's army flatly refused to face an ocean voyage

to a destination (as Dio says) 'outside the world', with the danger

of being wrecked at sea or cast away on a hostile shore. But

whereas in the case of Caesar's expedition the incident closed in

tragedy, in that of Claudius it turned into farce. One innova-

tion of Claudius's reign was the formation of a State secretariat

within the Imperial household, staffed by freedmen; and when

news reached Rome that the army of Britain was in mutiny, he

sent his chief secretary, the freedman Narcissus, to deal with the

mutineers. Roman soldiers were well accustomed to have an

imperator ascend the tribunal and address them; Claudius him-

self, as head of the army, they would have heard with respect;

but when an ex-slave mounted the platform they were indignant,

until their sense of humour overcame them, and they howled

the secretary down with shouts of 'Merry Christmas'—if that

will serve to translate the greeting that passed from mouth to

mouth at the Saturnalia, when slaves dressed up in their masters'

clothes and the social order was turned upside-down. Whether

because, having relieved their feelings, they thought better of

their obstinacy, or because Narcissus after an inauspicious begin-

ning handled his audience better than we are told, the mutiny

died there and then, and the legions went on board.

The fleet of transports sailed in three divisions, under separate

commands. It is safe to assume that there was a main body of

two legions, and two smaller divisions of one each. The motive

for this arrangement is recorded to have been the desire to avoid

being prevented from landing. Later in the history of Britain

the Romans made use of three ports on the coast of east and

south Kent, Richborough, Dover, and Lympne; and three

roads leading from these ports converge at Canterbury. It has

therefore been suggested that Plautius sent one flotilla to each

of these three ports, with orders to land and effect a concentra-

tion at Canterbury. There are two reasons for thinking this

unlikely. First, the strategy of the whole invasion was modelled

closely on that of Caesar's campaign; and, whereas Caesar had

explored the road from the east Kent coast to Canterbury, he had

left no information suggesting that good routes might be found

thither from Dover and Lympne, and had once for all settled

the unsuitability of Dover as a landing-place on a hostile shore.

Secondly, it is a maxim of strategy that forces should not be

80

THE CLAUDIAN INVASION

divided in the face of the enemy. Had the Britons prepared a

force at all equivalent to Plautius’s own for the purpose of oppos-

ing him at or after landing, and had any of his three divisions

encountered it on the beach or come into touch with it shortly

after disembarkation, his campaign would have opened with at

best a rebuff and at worst a disaster which would have been

poorly requited by a successful landing on the part of the other

two divisions.

It is more likely that when Plautius divided his forces ‘in order

that they might not be prevented from landing anywhere’1 his

intention was to confuse the Britons by making feints at two

other possible landing-places, of which Dover was very likely

one, while intending actually to land at one only. Recent

excavation has proved that this one was Richborough, where

the remains of a very large camp of Claudian date have been

found. Its landward defences form a crescent-shaped line, still

700 yards long, and originally longer, with both ends resting

upon the sea; inside the area so defended were many wooden

store-houses, of which the earliest were built soon after the land-

ing of Plautius’s army. The discovery of this large land-locked

harbour, hidden away behind the Isle of Thanet, where a great

fleet could ride at anchor or lie on the beach secure from danger

in any weather, was a triumph for the intelligence service of

Claudius’s army. It suggests that his staff realized where lay

the fundamental weakness of Caesar’s campaign, namely, in his

failure to find a safe and commodious harbour. By good fortune,

that weakness could be mended with a very slight modification

of his strategy, for the harbour of Richborough lay only a few

miles from his own landing-place.

In point of fact, Plautius’s precaution was unnecessary. The

Britons remembered the fiasco of Gaius; they had news of the

mutiny, which had lasted for several weeks; and they concluded

that there would be no invasion. Indeed, to speak of the Britons

collectively, as if they formed a single political unit, is mis-

leading. Britain at this time, although Cunobelinus is called

rex Britanniae by Suetonius, was very far from being one realm.

Cunobelinus, to judge from his coins, ruled not only over the

kingdom of Cassivellaunus but, as has been said in the preceding

chapter, Essex and Kent as well. Within this realm local auto-

nomies had been suppressed; but tribal identity and tribal con-

1 Dio, lx. 19, § 4.

sciousness remained, ready to assert themselves piecemeal when occasion offered. At the moment of the invasion, Cunobelinus was lately dead, and his sons' intention was probably to divide his inheritance between them; but this appears not yet to have been done, and actually we find two of the anti-Roman sons, Togodumnus and Caratacus, exercising a kind of joint rule which must have been a provisional and temporary arrangement, favourable to the invaders.

Outside the realm of Cunobelinus were many tribal states, some of which recognized the authority of Camulodunum, while others did not. There must have been a king reigning at Silchester, whether Epatticus or a successor of his, intimately related to the sons of Cunobelinus. Farther west, the Durotriges, a lately Belgicized tribe, doubtless had a king of their own in their stupendous fortress of Maiden Castle. The Belgicized inhabitants of west Sussex, the Regni, were ruled by Cogidubnus; the Iceni of East Anglia by Prasutagus; but alike in these cases and in those of tribes farther afield like the Dobuni, the Coritani, the Brigantes, and the Dumnonii, it is impossible to say how far the unity of tribal names and districts which has come down to us conceals a diversity of petty states, each with its own ruler, in the time of Claudius.1

In this state of things a unified policy for the defence of Britain was not to be expected. A Cunobelinus in his best days would no doubt have imposed such a policy, at least over the whole south-east; but Claudius rightly calculated that amid the family quarrels and party strife which marked his last years, and intensified after his death, there would be no organized resistance. Caught unprepared by Plautius's landing, the inhabitants of east Kent took to the woods, hoping to avoid a decisive battle and to wear the Romans out, as Cassivellaunus, according to what must have been their version of the story, had worn out Julius Caesar. Plautius spent the first stages of his campaign marching and countermarching in east Kent, searching for an enemy whom he never found.

In describing this phase of the campaign, Dio explains the impossibility of discovering and engaging the enemy by saying

1 Tacitus, Agr. 12, § 1, suggests that by the time of the Roman conquest a good many tribes had abolished the kingship and had a republican constitution. But his words must not be taken as implying that all had done so; that could be refuted from other statements of his own.

82

THE CLAUDIAN INVASION

that 'the Britons (of those parts) were not self-governing but were ruled by other kings'.1 If the tribes of Kent had been independent, each would have had a king of its own, an army of its own, and a capital of its own; Plautius could therefore have attacked them in detail, stormed their fortresses one by one, and forced them one by one into submission. As it was, he might overrun the territory of any one such tribe, only to find that it contained no army to fight, no fortress to attack, and no government upon which to impose terms of peace.

Meanwhile, however, the two most active sons of Cunobelinus were mustering their forces and hastening to meet the invaders. Even now, disunion was their undoing. Instead of acting in concert, each of the two brothers independently gathered his own men around him and rushed blindly upon the Roman force. Caratacus, the abler and more vigorous of the two, reached Kent first. It is tempting to conjecture that he took up his position on Caesar's old battlefield at the crossing of the Stour; for Dio's narrative suggests that he fought on ground of his own choosing somewhere in the eastern half of the county, and no better defensive ground is to be had. But Plautius found no difficulty in driving him headlong from his position. He escaped with his life and the remnants of his force along the line of the Watling Street; and Plautius in his pursuit, somewhere along that line, met with Togodumnus and crushed him. The loss of these two engagements made it impossible to hold East Kent, and some part at least of its inhabitants submitted to Plautius. They are described as a section of the Bodunni, who were doubtless one of the four unnamed Kentish tribes mentioned by Caesar.2

Rome's method of dealing with conquered tribes had undergone a change in the last hundred years. Caesar would have received hostages from the Bodunni; Plautius detached a small

1 Dio, lx. 20, § 1. The emendation 〈ἄλλοι〉 ἄλλους, which would imply that each tribe was ruled by its own king, makes nonsense of the passage.

2 At and before the conquest there were numerous tribes in Britain whose names are unknown to us, because the tribal names which have been preserved are mostly those of the cantons which which the Romans made into units of local self-government, and when this system was created many smaller tribes were merged in larger ones and their identity lost. It is therefore an error to assume that the name of a tribe mentioned in these earlier times will reappear later. Through overlooking this, Hübner (Römische Herrschaft in Westeuropa, 1890, ch. 1) was tempted to identify the Bodunni with the Dobuni and reduce the narrative of the conquest to a strategical nightmare.

force to police their territory, planted it in a little fort there, and

marched on. This way of disposing small police-forces in block-

houses can be traced back to Tiberius's northern campaigns in

the reign of Augustus;1 it was part of the process by which the

field-armies of the late Republic developed into the network of

sedentary garrisons that guarded the frontiers of the Empire.

These first engagements were mere skirmishes. While they

were going on, the general levy of fighting-men from north-west

Kent and from beyond the Thames had mustered on the Med-

way and encamped along the wooded hills west of the river,

where they were joined by the survivors from the defeated forces

of Caratacus and Togodumnus. It is probable that a bridge

already spanned the river at Rochester; if so, the retreating

Britons destroyed it and, imagining that this would give them

sufficient respite from attack, neglected to guard the crossing.

Plautius came on them unprepared. His Gallic cavalry swam

the stream and charged the British charioteers, with orders to

destroy their mobility by crippling their horses. Meanwhile the

future emperor Vespasian led the Second legion upstream to

find a ford, crossed the river, and turned the right wing of the

defenders. But the position was not to be so easily won. The

attack was beaten off. Next day battle was renewed on the

same ground, and it was only after severe hand-to-hand fight-

ing that Plautius gained the summit of the ridge, and looked

down to see the Britons retreating across the plain towards

Dartford.

The battle of the Medway was the decisive engagement of the

campaign. It was a hard-won victory for the Romans, for a two-

days' battle is a rare thing in ancient warfare; and it reflects

credit both on British leadership and on the steadiness of the

British troops that, though caught unawares and confronted

with a simultaneous cavalry attack and flanking movement, the

Britons were victors in the first day's fighting. It is clear that

their whole available forces were engaged, and that their plan

had been to put their main effort into the defence of the

Medway.

This fact is concealed from us in Dio's narrative, and for a

curious reason. From the first, the Roman plan of campaign

had involved the presence of Claudius at its culminating point.

1 Velleius Paterculus, ii. 120: after the defeat of Varus, Tiberius 'mittitur ad

Germanicam, Gallias confirmat, disponit exercitus, praesidia munit'.

84

THE CLAUDIAN INVASION

Plautius, a steady and experienced general, was to land the expeditionary force in Britain and take it through the early stages of the campaign; he was then to encounter the main forces of the enemy in a strong defensive position, and go through an elaborate comedy of professing his own incompetence to proceed farther and summoning the most unmilitary of emperors to take command and save the situation. Caesar's narrative of his own invasion, which was used throughout the campaign of 43 as a text-book by the Roman staff, would suggest to any reader that the right place for this picturesque event was the crossing of the Thames, where a great river guarded the frontier of Cunobelinus's immediate realm. Here, it was assumed, the Britons would make their chief stand and the decisive battle would be fought.

When Cassivellaunus took command against Caesar, he decided to fight him in east Kent. It was here that the battle took place after which, in Caesar's words, the Britons never again encountered him in a general action. The line of the Medway was never held against him, and consequently he never mentions its existence. Hence, when Plautius was marching across Kent in Caesar's tracks, he knew nothing of the Medway, and found himself quite unexpectedly confronted with the main army of the Britons and committed to fighting the decisive battle of the war. This explains his defeat on the first day. Plautius was fighting by the book, and on paper neither the Medway nor the Britons had any right to be there.

By comparison, the crossing of the Thames was a trifling affair. Relatively to sea-level, the land about London stood fifteen feet higher then than it does now. A trading settlement already existed at London itself, and the Thames had been bridged; but there were fords down-stream from it. The Britons in their retreat from the Medway naturally made for these fords, and crossed them with the Romans in pursuit. The Gaulish cavalry missed the fords, but some of them swam the river and others galloped for the bridge. The Britons made no attempt to hold the crossing. They continued their retreat towards Colchester. In the marshes of the Lea valley Plautius's cavalry came up with them; and now at last the Britons turned on their pursuers and checked them sharply.

There was nothing in this to cause the alarm which, according to Dio, Plautius professed to feel. A general with a free hand

CLAUDIUS ENTERS COLCHESTER

85

would have had his legions in support, and would have swept

the Britons before him until he reached the unwalled city of

Colchester. But Plautius was bound by the plan. If the Britons

would not defend the Thames, that was their affair; his business

was to make the most of his cavalry’s losses, send for the em-

peror, and hope that by the time he arrived the enemy would

have occupied the left bank in respectable force.

All went well. Claudius had reinforcements ready prepared,

including a contingent of his own praetorian guards and an

elephant corps; he made a rapid journey to Britain, arrived in

Plautius’s camp, and gave orders for a general advance. At

this point our sources become contradictory. Dio says that the

crossing of the Thames was opposed and a battle fought in

which the emperor was victorious; Suetonius, that no engage-

ment took place and no casualties were incurred during his

whole stay in Britain; Claudius himself, in the inscription cut

on the triumphal arch celebrating his conquest, that he suffered

no losses.1 If there was opposition, it must have been of the

slightest; the Britons had shot their bolt at the Medway, and

had no heart for another battle.

At Colchester Claudius received the submission of many

tribes2 and laid down the lines for a peaceful settlement of

Britain. His biographers have noticed that alike in his adminis-

tration and in his reforms he presents a curious mixture of

bold innovation with an almost pedantic conservatism. The

systems of government which he created, whether at home or in

the provinces, never have the simple clarity of a scheme deduced

from logical principles. They are patchwork affairs, composed

of heterogeneous and conflicting elements. In this, his British

settlement is characteristic of the man.

To begin with, there was the realm of Cunobelinus. This was

the part of Britain which had directly incurred the hostility of

Rome, and the occupation of its capital marked the close of the

first campaign and was the occasion of the emperor’s triumph.

Of its royal princes Togodumnus was dead; Caratacus was a

1 Dio, lx. 21, § 4; Suetonius, Divus Claudius, 17, § 2; Dessau, Inscr. Lat. Sel. 216.

Perhaps the inscription (sine ulla iactura) is the source from which Suetonius gets

his sine ullo proelio aut sanguine.

2 Eleven kings submitted to him, according to the inscription already quoted;

but that inscription was only put up in 51, and the eleven may include all who

submitted down to that time. If so, however, the phrase sine ulla iactura is unjusti-

fied: it can truthfully refer only to the sixteen days of the emperor’s own visit to

Britain.

3720.1

G

86

THE CLAUDIAN INVASION

defeated and discredited exile. Here Claudius's course was

obvious: this kingdom was to be a Roman province, with Col-

chester as its capital and Plautius as its first governor. But the

straggling, half-barbari town of Cunobelinus was no fit capital

for a province. A new city was laid out on the hill-top a little

way to the south-east: here in all probability was the official

residence of the propraetor and the seat of Roman government;

and here in the centre of the city, surrounded by the porticoes

of a forum, was to stand a temple, a vast and massive building

where Claudius was to be worshipped as a god.

Claudius was ready to give with one hand what he had taken

away with the other. The Belgic Britons had shown themselves

willing to adopt town life of a sort and able to conduct their

own public affairs; these signs of grace, then, should be fostered.

In his other new-conquered province of Mauretania, Claudius

had converted at least one native town into a self-governing

municipium; the same experiment should be made in Britain, and

for this experiment no place could be more suitable than the

Catuvellaunian capital where Tasciovanus had once rcigned.

Accordingly, a new city was built at Verulamium, in the valley

below the old town, and endowed with municipal rights.¹

Outside the realm of Cunobelinus there were at least two

independent states which had taken no part in the war and

now lost no time in submitting to the conqueror. One was the

Regni of west Sussex. We have already seen that this was a

region cut off from the Belgae of the lower Thames basin by

the forest of the Weald, but affected to some extent by the La

Tène civilization of the south-west and more decisively, later

on, by the western Belgic invasion, under whose influence its

people had ceased to inhabit their hill-forts and had built them-

selves a new capital at Noviomagus in the plain. Here, at

Chichester, ruled a king named Cogidubnus. We may assume

that he was no friend to Caratacus and Togodumnus; and so

convincingly did he make his peace with Claudius that the

emperor confirmed him in his kingdom and even enlarged its

boundaries, conferring upon him a title thoroughly charac-

teristic of his tendency to mingle innovation with conservatism:

rex (et) legatus Augusti in Britannia. Strange though the title is,

its spirit is not out of harmony with the methods of Claudius.

In the following year he revived the title of rex for the benefit

¹ If so, it lost those rights by the early third century (cf. C.I.L. vii. 863).

TIBERIUS CLAUDIUS COGIDUBNUS

87

of M. Julius Cottius, representative of the ancient native rulers of

the Cottian Alps; his father had enjoyed only the title of praefectus

civitatiun, but his grandfather Donnus had been a king. Simi-

larly in Judaea Claudius experimented from time to time in the

re-establishment of client kingdoms, taking care, however, that

the men whom he elevated to these petty thrones should refrain

from striking out independent policies of their own.

Cogidubnus accordingly was obliged to give proofs of loyalty.

He became a Roman citizen with the name of Tiberius Claudius

Cogidubnus, and either he or some wealthy subject acting

with his authority employed Roman artists to build a temple

to Neptune and Minerva, dedicated for the welfare of the im-

perial house, whose dedicatory inscription, the most elegant and

purely classical in Britain, still survives at Chichester to tell the

story. The choice of deities is not without significance. Chiches-

ter stands at the head of the first and most easily accessible of

those Hampshire harbours which, even as early as Strabo's

time, were connected by regular trade with the mouth of the

Seine; and the god of the sea, named together with the goddess

of learning and the arts, suggests that Cogidubnus in building

this temple was publicly announcing his attachment to the new

civilizing influences which cross-Channel trade was bringing to his

coast. How far his title of imperial governor in Britain indicated

a real authority over his own people, and how far it was merely

honorary, serving to cover a real subordination to Plautius, it

is difficult to say, but the later history of Prasutagus suggests that

his authority was genuine and not lightly to be revoked.

The second of these client kingdoms was that of the Iceni.

This tribe had always held themselves aloof from their Belgic

neighbours, and in all probability lived in daily fear of them.

Eighteen years later they still had a king of their own, Prasutagus;

either he or his father was presumably on the throne at the time

of the conquest, and shared with Cogidubnus his submission

to the emperor and the emperor's gift of a partial and patronized

freedom. A third client kingdom ought probably to be added;

that of the Brigantes.

Having conjured up this strangely devised political system for

his new province, Claudius instructed his propraetor to conquer

the rest of Britain, and went home to the devout enjoyment of

his triumph.

VI

CARATACUS AND BOUDICCA

AFTER the battle of the Medway Caratacus vanished. He knew that it was decisive, and that all further attempts to

defend his kingdom and capital must be futile. But he was a man of indomitable spirit, whose power over others rested not

upon his royalty but upon his personality. He resolved to gather round him a small band of kinsmen and friends, to cut himself

loose from his own country, and to rouse in the tribes as yet un-touched by Rome a spirit of resistance to the invaders.

It was a bold step for the king of the Belgic Britons to take. Both he and his people were rather feared than loved by their

neighbours. Their conquests were certainly resented, and their civilized manners were no doubt despised; the other tribes were

likely, on the whole, to be pleased at the swiftness and completeness of their humiliation. The Regni and the Iceni, and many

other neighbouring peoples, had long hated them, and had only waited to show their hand until they were no longer to be feared.

Where, then, was Caratacus to go?

With the strategical instinct of a born soldier, he saw that the highland zone was the place in which the legions must be re-

sisted, if they were to be resisted anywhere. He saw that, of its three sectors, the Dumnonian peninsula was too small and too

much isolated. He saw that it was only in Wales and the north that he could hope for success. The north was already a great

military power under the rule of the Brigantes; in Wales, hardly touched as it yet was by Iron Age civilization, there were the

materials for another such power, lying ready to the hand of any one who could use them: a hardy and independent race,

and a country designed by nature for desperate defence. He turned to Wales.

It must be asked, even though the question cannot be con-clusively answered, why he did not turn first to Wessex, where

the former realm of Commius offered him a people akin to his own, and where, as he travelled westward, he could find a

country rich in fortresses, each one capable of stout defence. A possible answer is to be found in the defection of the partially

Belgicized Regni. If Cogidubnus was not to be trusted, the same may have been true of the other chiefs in central south England.

CARATACUS IN THE WEST

89

Jealousies between them and the dynasty of Cassivellaunus may have been too strong for Caratacus to expect their sympathy.

And he was no diplomat. He was a man of fiery and over-bearing temper, totally unsuited to the work of begging help

from princes who thought themselves his equals and looked on him as a discredited exile. The later events of his life showed

that his haughty bearing could not gain the support of a Cartimandua, though it could win the respect of a Claudius. Such

a man, driven from his own kingdom, had the best chance of success if he tried to carve out a new kingdom for himself among

the masterless tribes of the west.

His temperament as a fighter helped this decision. He was not the man to conduct a dogged rearguard action, to dispute

every step of an advancing enemy and to wear him out by constant pressure on his flanks and his communications. The course

of the campaign of 43 after the battle of the Medway showed that Caratacus never had it in him to emulate Cassivellaunus’s

tactics against Caesar. He was a fighter of battles; his idea of warfare revolved round the conception of the decisive engage-

ment. His first campaign had culminated in one such engagement; he had lost it; and he now set to work to prepare an army

and a field of operations for a second.

Meanwhile Plautius was proceeding systematically with the conquest of the lowland zone. It is a process of which ancient

writers tell us practically nothing; and archaeology has done very little to fill the gap. Beside the great camp at Richborough,

there is hardly a single place where we can identify one of the marching-camps or forts which during these first years must

have been scattered broadcast over the face of the country. But the few fragments of information which have reached our hands

make a picture which, so far as it goes, is intelligible.

The Roman supreme command made the assumption that henceforth no great concentration of enemy forces would be

encountered. Their own forces could therefore be divided into separate columns, each of which was to work in one direction

and thus, by a general radial movement, overrun and pacify the lowlands. Sussex and East Anglia were friendly; therefore the

area to be conquered spread fanwise from the Thames estuary westward and northward. The centre of the advance would

run through Catuvellaunian territory into the forests of the midlands; the right wing would work northwards towards

90

CARATACUS AND BOUDICCA

Lincolnshire, the left would overrun Wessex from east to

west.

Although Colchester was the capital of the new province, the

centre from which these radiating lines diverged was at London.

The building of roads was an essential instrument of conquest,

and most of the main Roman roads in Britain must have been

laid down at a very early date. When we find London the centre

from which one road, Watling Street, runs north-westward to

Wroxeter, another northwards to Lincoln, and a third westward

to Silchester, we can hardly doubt that these were the roads

built for this triple advance, and that London served as the

supply-depot and base for the three columns.

Of the central column's fortunes we know nothing. It pro-

bably consisted of the Fourteenth and Twentieth legions; and

it pushed its road forward along the central watershed of Eng-

land, crossing the rivers close to their sources, gradually leaving

the Catuvellaunian tilled lands behind and plunging into the

woodland of Warwickshire. On its way it encountered the La

Tène civilization of the Northamptonshire uplands, but after

this its march lay through country almost uninhabited. We may

wonder why so powerful a column was sent into this desolate

and unpeopled land; the answer must be that its destination was

already fixed on the Welsh borders and the Irish sea, and that

its ultimate function was to conquer Wales before turning, if

that should be necessary, against the Brigantes.

The right wing consisted of the Ninth legion. Its route lay

through Cambridgeshire and round the edge of the Fens; but

once more, we know nothing of its operations, unless the

Roman camp lately seen from an aeroplane near Castor1 is

a relic of its progress. Unless they had already submitted to

Claudius, the Brigantes must now have made peace; or rather,

their queen Cartimandua did so, now if not earlier, but was

unable (as we shall see) permanently to impose her will upon

all her subjects.

About the left wing we know more, because it was com-

manded by the future emperor Vespasian, whose biography

tells us that in Britain he 'fought thirty battles, conquered two

powerful tribes, capturing over twenty fortresses, and annexed

the Isle of Wight'.2 He commanded the Second legion, which

1 Antiquity, iv (1930), pp. 274-5.

2 Suetonius, Vespasian, 4.

has left a trace of its presence on the south Devon coast. What

the two tribes were it is idle to guess. The Atrebates of Sil-

chester may have been one, or they may have submitted of their

own free will; one must almost certainly have been the Duro-

triges, whose fortress of Maiden Castle was surely one of the

twenty-odd that Vespasian stormed. He may have overrun the

country of the Dumnonii; there are earthworks in Cornwall

which may possibly be the camps of a Roman army.1

By the autumn of 47, when Plautius was succeeded as gover-

nor by P. Ostorius Scapula, the right wing had reached Lincoln,

the left was on the borders of Devonshire and Dorset, and the

centre was somewhere in the midlands. But the situation was

far from satisfactory. Ostorius found the friendly tribes much

harassed by constant raiding from those which were still un-

conquered, especially the Brigantes and Silures, the Brigantes

crossing the Trent and raiding the Coritani of Leicestershire

and Nottinghamshire, the Silures crossing the Severn and plun-

dering the Dobuni. This counter-offensive was perhaps the

firstfruits of Caratacus's activity in the west. Ostorius either

knew or suspected that these raids were encouraged by unruly

elements in the tribes that had submitted. Accordingly, after

a series of sharp engagements in which the raiding forces were

repulsed by his auxiliary infantry and cavalry, he drew a

frontier-line across Britain, disarmed all the tribes on his own

side of it, and fortified and patrolled it to keep out raids from

beyond.

This frontier-line was the road which is known as the Fosse.

It was meant to hold down the entire country up to the Trent

and Severn; that is to say, it was planned so as to defend the

Dobuni and Coritani against marauders from across those rivers.

The way in which the Fosse is laid out shows that its designers

regarded the Dobuni and Coritani as friendly tribes, not as

enemies; it passed right through the centre of their territories

and was, according to the ordinary practice of the time, studded

with auxiliary units, whose duty was to patrol the neighbour-

hood of their various forts, and thus act as a defensive garrison

for the country in which they were placed, and upon which they

depended for supplies.

The Fosse begins at Seaton, by the mouth of the Axe on the

Devonshire coast, where a tile of the Second legion indicates

1 Vict. Co. Hist. Cornwall, i. 26, note 74.

92

CARATACUS AND BOUDICCA

that it once lay there in more or less permanent quarters; per-

haps a tradition of this residence is preserved in the pages of

Ptolemy.1 It runs, with a directness rare even in Roman roads,

to Bath and across the Cotswold plateau to Cirencester, which

either was or soon became the capital of the Dobuni; here

early tombstones suggest a cavalry garrison. About Moreton-in-

Marsh, still following its straight line, it gradually leaves the

plateau and enters the midland forest, and somewhere in this

region, perhaps near High Cross, north of Rugby, where it en-

counters Watling Street, the Fourteenth and Twentieth legions

must have had their encampments. At Leicester, a tile suggests

that a detachment of the Eighth, possibly part of Claudius's

reinforcements, garrisoned a fort where later, if not already, the

tribal capital of the Coritani stood. The road then follows a

ridge of hills on the right bank of the Trent to Newark; on this

ridge a Claudian fort has been excavated at Margidunum, east

of Nottingham. From Newark it runs straight across the plain

to the Ninth legion's fortress at Lincoln. The two ends of this

line, as well as its centre, where no river offers an impediment

to the free crossing of it, were thus already well defended by

legionary fortresses; it was the gaps between these points that

were exposed to Silurian and Brigantian raids, respectively, and

it was to protect these gaps that Ostorius planned his frontier-

line.

But his disarmament of the tribes lying behind the line was

unexpected, and in one quarter at least bitterly resented. The

Iceni regarded themselves as free allies of the Roman people,

and although they had hitherto paid their allotted tribute and

provided their quota of recruits without protest, they were too

proud and independent in spirit to submit to disarmament.

They rose in revolt and carried with them a number of neigh-

bouring tribes; certainly not the Brigantes, who were outside

the frontier, nor the Catuvellauni or Trinovantes, who were

under the direct rule of the propraetor, but small tribal units

whose names have not come down to us. Realizing from the

first that defence and not attack must be their part, they as-

sembled their forces in a strong position protected by field-works

1 Ptolemy, writing his Geography in the second century, places the Second legion

near the town of Isca, and calls this a town of the Dumnonii. That should be

Exeter; but Bradley (Collected Papers, p. 72) pointed out that he confused the Exe

with the Axe. Of course, a further confusion with the Monmouthshire Usk is con-

ceivable, but not probable.

and abattis. Scapula’s legions were busy constructing and fortifying the Fosse, and he could only spare auxiliaries to deal with

the revolt; but he dismounted his cavalry, since the position was one which only infantry could attack, and with these and his

infantry cohorts assaulted and stormed the rebels’ defences. In spite of desperate resistance, the victory was complete. The

Iceni and their allies were disarmed; but it is noteworthy that, in spite of the revolt, they were still allowed to retain the status

of a nominally independent kingdom.

Ostorius now thought himself free to move onwards, and his next campaign was on the northern frontiers of Wales in the

country of the Degeangli. It was now, presumably, that Watling Street was driven forward from High Cross to the upper

Severn, and a legionary fortress established at Wroxeter. Excavations have failed to discover any remains of such a fortress,

but early tombstones of men from the Fourteenth and Twentieth legions have been found there, one at least of which can

hardly have been set up later than the time of Claudius; and the fact that the forum of the later city was designed on the

model of a praetorium suitable for a fortress containing two legions1 gives further support to the suggestion that in or about

the year 48 a double fortress was established there, in which both those legions took up their quarters. It was from this new

base that Ostorius attacked the Degeangli, whose lands he overran without fighting a pitched battle.

He hoped in the course of this campaign to establish himself on the Irish sea, make contact with his fleet, and effect the

separation of Wales from Brigantia. But before he had reached his goal he learnt of trouble among the Brigantes which de-

manded his intervention. They were a turbulent race. Though Cartimandua had made peace with the Romans, that did not

prevent a considerable part of her subjects from treating them as enemies and plundering the lands of tribes whose submission

had been more complete. These were the raids which had given trouble to Ostorius at the beginning of his term of office. His

suppression of them, no doubt accompanied by a demand that Cartimandua should in future keep a firmer hand on her sub-

jects, caused the more insubordinate elements among the Brigantes to turn against the queen herself, and it was this civil war

1 This was pointed out by Professor D. Atkinson, excavator of the forum. His report has not yet been published; but see Macdonald, Roman Britain 1914-1928.

94

CARATACUS AND BOUDICCA

which distracted Ostorius from his campaign in Flintshire. He

marched into Brigantia so promptly that he was able to nip the

revolt in the bud, punish those who had committed themselves,

and pardon the rest.

If the Silures had been as tractable as the Brigantes, Ostorius

would soon have been able to resume his forward movement.

But conditions here were very different. The pressure on his

left wing, directed against the long and lightly held Gloucester-

shire sector of the Fosse, was unremitting. Punitive expeditions

did not check it; negotiations were fruitless. The fact is that

Ostorius was here confronted, not by the turbulent subjects of

a friendly or submissive ruler, but by an organized guerrilla

campaign, whose brain was Caratacus. Since his defeat on the

Medway and his flight to the west Caratacus had done much to

restore his reputation. Even before the arrival of Ostorius in 47

he had gained an ascendancy over the Welsh tribes and begun

to lead them against the conquered west in a series of systematic

raids. For the last two or three years (we are now speaking of

events happening about the year 50) he had continued on the

same course, achieving a substantial degree of success in spite

of the garrisons in the forts of the Fosse. It was, in fact, the

standing weakness of the Roman frontier system, by which small

police-posts were scattered in a chain or network over the belt

of country forming the frontier, that forces so minutely sub-

divided were helpless against a mobile and concentrated enemy

whose movements were directed by an able soldier. In such

warfare the initiative lay with Caratacus, who was the right

man to use it; and he was now worshipped by his followers as

a national hero, the champion of their liberty, and their heaven-

sent leader against the invaders.

The only way in which Ostorius could recover the initiative

was by placing a strong force in the lower Severn valley, in

advance of the Fosse frontier, where it could take the offensive

against the Silures and break up any concentration in their

country before it became a danger to the country of the Dobuni.

This implied the creation of a legionary fortress on the left bank

of the Severn at some point where it could be easily crossed,

threatening the hostile right bank somewhat as Cologne or

Mainz threatened the right bank of the Rhine. The obvious

position for such a fortress was Gloucester, which is still the

lowest bridge on the Severn; and here Ostorius quartered the

THE LAST STAND OF CARATACUS

95

Second legion, bringing it up from the south where its work of pacifying the Dumnonii was now complete.1

But Ostorius had not forgotten the Icenian revolt, and before he undertook his final campaign against Caratacus he decided to safeguard the peace and loyalty of the south-eastern Britons by planting a colonia at Colchester, in order that a strong body of time-expired legionaries, living on their allotments of land and ready to take up arms again when occasion demanded, should serve as a garrison for the capital and a threat to neighbouring disaffection.

This required that the Britons of the surrounding country should be dispossessed of their land after having believed that their tenure of it was secure; but Ostorius was not a conciliatory governor, and cared little for the feelings of natives when a military question was in the balance.

The event showed that his judgement was at fault; the resentment caused by this expropriation created more disaffection than the presence of the colonists could quell.

But it was another ten years before that harvest was ready for reaping.

In 51 all was ready for the final blow.

The legion at Gloucester, planted there probably a year earlier, had already destroyed Caratacus's freedom of action among the Silures, and he had moved into central and north-western Wales, the country of the Ordovices.

When he knew that Ostorius was coming in search of him, he chose his ground with care for a decisive engagement.

The place which he chose resembled Caesar's battle-field at Bigbury, transposed into terms of a mountain country.

A craggy hill, its summit ringed by a stone-walled fort, overlooked a river; the fort was the centre of the British position, and the rocky slopes between it and the river were to be the scene of the first encounter.

Caratacus had learned that a legion in line of battle was practically invincible to the British warriors, with their fluid and shifting formations; but he calculated that no troops, however well drilled, could keep shoulder to shoulder while advancing over such steep and broken ground, and that his skirmishers stood a good chance of repulsing every Roman charge before

1 There is no proof that the Second or any other legion ever had a fortress at Gloucester.

But Tacitus's statement that Ostorius placed a legionary fortress in the country of the Silures, or at least on its borders, cannot refer to Caerleon-on-Usk, which was not founded until about 75; and not only is Gloucester the natural place for it, but the shape and size of the Roman town there are exactly those of a legionary fortress, and suggest that this may have been its character in the first phase of its history.

96

CARATACUS AND BOUDICCA

it reached the walls of the fort where his last stand was to be made.

To Ostorius, though his total force, legions and auxiliaries together, far outnumbered that of Caratacus, the British position seemed wellnigh impregnable; and the skill and ardour with which Caratacus was preparing his defences, placing his men, and encouraging every unit to do its best, made the Roman general, as he watched him, realize the mettle and quality of his antagonist. But his men were in high spirits, and their officers keen to lead them; and Ostorius, knowing that it would be a soldiers’ battle where everything must depend on the spirit of the troops, resolved to fight. He divided his forces into several columns and launched them simultaneously against different points where the river seemed fordable and the hill easy of access. Their first rush carried them across the stream and far up the slope. At the walls of the fort they were brought to a standstill, galled by a constant fire of missiles from above; but they locked their shields overhead, tore down the stonework, and broke in on every side. The Britons retreated, fighting stubbornly, to the very summit of the hill; here, surrounded and hopelessly outnumbered, the remnant of the defenders at last yielded. Among the prisoners were the wife and daughter of Caratacus, and his brothers, but Caratacus himself had, for the third time, fled from the scene of his defeat and escaped to renew the war elsewhere.

This time his hope was vain. In Wales he had shot his bolt; there remained Brigantia. His only chance was to persuade Cartimandua to abandon her position of a client ruler and raise the standard of rebellion. But Cartimandua was deeply committed to Rome. Not only had she made submission, she had called upon Roman aid against her own rebellious subjects, and knew that if she made Rome her enemy her own throne was forfeit. She threw Caratacus in chains and sent him to Ostorius as a pledge of her loyalty; and the last son of Cunobelinus, after facing with dignity the emperor and populace assembled to triumph over him, ended his days in honourable captivity at Rome.

In Britain, now that the assembled levies of Wales were routed and the fidelity of Brigantia proved, it seemed that the war was over. No more operations in the field were undertaken. Ostorius began to build forts in the country of the Silures, treating

FURTHER TROUBLE WITH THE SILURES

97

them as a conquered tribe that only needed now to be kept quiet.

But the Silures were only awaiting an opportunity to avenge

their hero; and the dispersal of the Second legion in working-

parties over the country-side gave them their chance. At a con-

certed moment the whole tribe rose. The half-built forts were

attacked and destroyed; the legion’s praefectus castrorum and eight

centurions, with many rank and file, fell in bringing support

to the hard-pressed working-gangs; and although in spite of

these losses the legion beat off the Silurian attack, the assailants

did not give in. They hung about the fortress, and watched for

an occasion to strike a blow. Once they overwhelmed a foraging-

party and cut up the cavalry sent out to rescue them; routed the

auxiliary infantry sent in support of the cavalry; and were only

put to flight when Ostorius, now present in person, led out his

legions, when they escaped into the gathering dusk. This was

only one of many engagements in which the Silures, by swift

and well-timed strokes, inflicted heavy loss on the Romans and

suffered little in return, sometimes carrying off prisoners and

plunder, sometimes merely destroying, but always making good

their escape before the counter-blow could be delivered. Their

spirit was only heightened by learning that the Roman governor

had sworn to exterminate the entire tribe; and their example,

backed by gifts of plunder and prisoners, had already begun

to have its effect upon other tribes hitherto submissive, when

Ostorius, worn out with fruitless work and disappointments,

died.

Before his successor, Aulus Didius Gallus, could arrive matters

had gone from bad to worse. The Second legion had been

beaten in the field, and the Silures were carrying destruction

far and wide into Roman territory. The first task of Didius was

to recover lost ground, clear his subjects’ land of Silurian raiding-

parties, and re-establish a state of defence in the Severn valley.

No sooner had he done this than the Brigantes began to be

troublesome. They were the most important of the tribes that

had been worked upon by the recent Silurian successes. Carti-

mandua stood obstinately faithful to Rome; but Venutius,

hitherto the partner both of her throne and of her policy, now

put himself at the head of a faction to dethrone the queen and

turn the whole power of Brigantia against the Romans. Didius

hurriedly dispatched a force of auxiliaries, by whose help she

succeeded, not without difficulty, in maintaining her position

98

CARATACUS AND BOUDICCA

against the rebels; a legion followed, which eventually put down

the rising. Thus for the second time Cartimandua was con-

firmed in her rule by the armed intervention of a Roman

propraetor.

Didius governed Britain from about 52 until 58. An elderly

and unenterprising man, he was content to protect his province

against dangers threatening it from Silurian aggression and

Brigantian turbulence without advancing its frontiers; his only

forward step consisted in the establishment of a few auxiliary

forts, no doubt in the country of the Silures.1

The next governor, D. Veranius Nepos, was ambitious and

energetic. He came out to Britain resolved to make his mark.

He inherited a long tradition of ill success against the Silures,

who were quict but far from cowed; and he determined to settle

accounts with them promptly. His premature death, before he

had held office for a year, put an end to the insane hope,

revealed when his will was read, of completing the con-

quest of Britain within two years. In 59 he was succeeded by

C. Suetonius Paulinus, one of the most distinguished soldiers

of the day.

Paulinus's first two years are said to have been marked by

conquests and by the establishment of new auxiliary forts. It

is possible to infer the direction which this forward movement

took. The Brigantes were quiet again under the submissive

Cartimandua. The Silures were not finally conquered until half

a generation later. It is clear, therefore, that Paulinus, satisfied

with the position established on these two flanks by Didius, took

up the project of an advance to the Irish sea where some ten

years earlier Ostorius had been compelled to lay it down. He

had at his disposal the Fourteenth and Twentieth legions, lying

at Wroxeter; with these he repeated Ostorius's march through

the country of the Degeangli, building forts as he went, and

penetrated steadily and deeply into North Wales.

By 61 he had reached the Menai Straits, and in that year he

proposed to conquer Anglesey. He built transports for his in-

fantry; his cavalry were to swim or ford the shallows. His

crossing was opposed by a strange and awe-inspiring assembly.

1 Tacitus, Agric. xiv, § 3. We do not know how far the territory of the Silures

extended eastward; I assume that the Severn was the boundary between them and

the Dobuni. In any case, the plain on the left bank of the river was almost un-

inhabited.

DEATH OF PRASUTAGUS

99

Among the groups of warriors women in ceremonial dress, bearing lighted torches, ran hither and thither; behind them were druids, standing by the fires of human sacrifice, their hands raised in prayer for help against the invaders. The Romans at first shrank from so weird a sight; then, recollecting themselves, leapt ashore, fell upon warriors, priests, and women indiscriminately, cut them down, and flung them into their own fires. It was not a battle but a massacre. Beyond lay the fertile plain of Anglesey; but before Paulinus had overrun the whole of it he had news which turned him headlong back.

Prasutagus, king of the Iceni, had submitted to Claudius in the year of the conquest without striking a blow, and had been allowed to retain his throne as a client king. As we have seen, even the resistance of the Iceni to the disarmament ordered by Ostorius did not lose them this status. But in 61 Prasutagus died; and, whatever may have been the intention of Claudius, the policy of Nero was unfavourable to the continuance of client kingdoms; in Pontus and in the Alps, on the death or abdication of their rulers, he abolished them and established in their place the standard type of provincial government. Prasutagus left a widow, Boudicca, and two daughters, and by the old trick of making the emperor a coheir he had attempted to secure his children in their inheritance. But no sooner was he dead than a tempest of calamities burst upon his country. The royal line was treated as extinct and its property confiscated. The emperor's procurator, Decianus Catus, extended the confiscation to the property of all the Icenian nobles. The procurator's servants fell upon the palace, and in the sack Boudicca was flogged and the princesses raped. Nobles and kinsmen of the dead king were seized as slaves. The debts which had been incurred in the process of romanizing the upper classes of Icenian society—a process involving vast purchases from the traders who swarmed over the country in the wake of the legions—were hurriedly called in by panic-stricken financiers, chief among whom was the apostle of mercy and the simple life, the philosopher-statesman Seneca. And in the midst of all this ruin and outrage, the officers of the Roman government appeared, reiterating their twin demands for recruits and tribute-money.

Furious at this breach of faith, terrified at the prospect of endless future oppression, and burning to avenge the insult to their royal house and nobility, the Iceni rose at Boudicca's call. The

100

CARATACUS AND BOUDICCA

rising spread to the Trinovantes, who had neither forgotten nor forgiven the confiscation of their land by Ostorius when he founded the colony of Colchester. The rebels resented not only the ordinary levies and taxes of the empire but, still more, the special demands upon them made for the upkeep and service of Claudius's temple, the symbol of Rome's dominance among them.

And now the price had to be paid for Ostorius's neglect of administrative detail in his thirst for military success. The disarmament which he had ordered had been a farce: weapons were in every man's hand. The colony which he had founded had neglected the most elementary precautions of defence: it was an open town, where unarmed men retired from the legions lived at their ease. Too late they heard the mutterings of the coming storm. Stories were passed from mouth to mouth, how the buildings of the colony had been seen upside down in a mirage at sea; how the tide had flowed red with blood and had left on the beach things like heaps of human bodies; how strange cries had been heard in council-house and theatre. The statue of Victory fell from its base. Women screamed suddenly in the streets.

Paulinus was 250 miles away. There was no legion within 120. The procurator sent 200 half-armed men; the citizens could muster a handful more. When Boudicca's warriors appeared, on every side at once, there was no thought of defence. The few who had weapons fled to the temple, whose massive stone base was the only thing that could resist sword and torch. The rest of the town went up in flames. When the fire died down nothing was left except the blackened temple and its few score defenders, and in two days that was empty.

At Lincoln the Ninth legion was commanded by a gallant soldier, Quintus Petillius Cerialis. As soon as he heard that the Iceni had risen, he marched to the rescue of Colchester with a vexillation of two thousand men. Before he reached it all was over, and Boudicca had turned to meet him. Swamped by sheer numbers, the legionaries went down fighting. Only the cavalry, with Petillius himself, made good their escape, and fled closely pursued to the very gates of their fortress, where they shut themselves in and beat off their pursuers.

One other Roman escaped whom Rome could well have spared. Decianus Catus, the procurator whose rapacity and

BOUDICCA'S RISING

101

violence had been the chief cause of the disaster, did not stay

to share the consequences of his own action. He found a ship,

and crossed over to Gaul.

This was the news that reached Suetonius as he lay encamped

in Anglesey. He had with him, beside auxiliaries, the Four-

teenth legion and a part of the Twentieth; the rest was guarding

the fortress at Wroxeter or building roads and forts in North

Wales. His first thought was for his depot and base of supplies

at London; but that was 230 miles away, hardly to be reached

even by the best infantry in a week of forced marches, and in a

week anything might happen. He made a bold decision. Leav-

ing his infantry to follow, and sending a galloper to summon the

Second legion from Gloucester, he rode for London at full speed

with his cavalry. The country was ablaze; at any moment he

might encounter Boudicca's host and be engulfed; but he rode

on, and reached London in safety.

London was a large straggling town, unwalled, full of peace-

ful traders and military stores. It was utterly incapable of de-

fending itself. But the supply depot was no doubt, according to

custom, entrenched and palisaded, and held by a handful of

troops. If the Second legion arrived in time (and three or four

days' marching should bring it) the non-combatants could be

herded within the ramparts, and Suetonius, with one legion, his

own cavalry, the depot troops, and such citizens as he could arm,

had at least a chance of beating off the Britons until his main

body should come up. There were unknown factors as well.

After chasing Petillius to Lincoln, Boudicca might waste time in

storming auxiliary forts on the line of the Fosse and elsewhere;1

in that case his main body would reach London before her.

He had not been more than a day or two in London before

he knew that his hope of defending it was vain. The Britons

were hastening southwards. London was clearly their objective.

And, worst of all, the Second legion failed him. At the moment,

no legate was in command; and the officer in charge, the prefect

of the camp, lost his nerve and refused to face the risk of another

disaster like that which had overtaken the Ninth. For Suetonius

to remain in London, with the force at his disposal, would be

suicide. His duty was clear. He must rejoin his main body, and

1 When he wrote the Agricola (xvi, § 1) Tacitus was under the impression that

Boudicca had actually done this; later, in the Annals (xiv. 33, omissis castellis), he

corrects his earlier statement.

3720.1

H

102

CARATACUS AND BOUDICCA

choose time and place for a defensive battle in which, despite the

enormous odds against him, there should be a bare possibility of

success. He hardened his heart to the prayers of those whom he

was leaving to certain death, and taking with him all who could

march, set off again up Watling Street.

At Verulam, still less defensible than London, the same ordeal

awaited him. He pressed on, and somewhere in the midland

forests he met his troops and made his preparations for a last

stand. Meanwhile Boudicca reached London. There was no

resistance, and no prisoners were taken. To this day, men dig-

ging in the city find everywhere the layer of ashes which is all

that was left when her men had done their work.1 Then they

turned to Verulam, and that went too. In those three towns it

was said that 70,000 perished, by sword or gibbet, on the cross

or in the fire.

Suetonius had only 10,000 men to meet a triumphant and

confident enemy that far outnumbered him. But he knew that

their success would make them reckless, and that if he could

avoid being surrounded and keep his men steady all might be

well. He chose a battle-field where flank and rear were pro-

tected by dense woods, so that the attack must be delivered

square on his front; and placing his cavalry where they could

strike at the flanks of his assailants, he awaited the event. The

Britons, an unwieldy host accompanied by their wives and

children in wagons, only cared to trap him on his own ground;

they drew up their wagons so as to close the only outlet, and

came on to offer battle in front of them. Suetonius did not wait

until they were ready to rush him. He charged, foot and horse

together, and swept them in confusion against their own wagon-

barrier. The surprise was complete, and escape impossible.

Men and women, even the wagon-horses, were cut down until

the narrow front was choked with the dead. The British host

was wiped out as completely as the towns it had destroyed.

Boudicca took poison; when the news reached Gloucester, the

prefect of the Second fell on his sword.

The revolt was over, but there was much still to do. Reinforce-

ments were sent from Germany and the Ninth legion brought

1 There is a place in Spitalfields where the bones have been found of people who

appear to have been buried by cartloads. Archaeological evidence of date is lack-

ing; but their skulls are like those of Italians of the same time, and it is possible

that they were some of the Roman settlers who perished in Boudicca's massacre

(Morant, Biometrika, xxiii (1931); Buxton, F.R.S. xxv (1935), p. 40).

up to strength; and now Suetonius took the field for vengeance. New police-posts were scattered over the country, and the land of the guilty and suspect tribes was ravaged with terrible thoroughness. As the year wore on, famine helped. The rebel-lion had begun in the spring, and fields had remained unsown, partly because every man had joined the rebels, partly because they had counted on using the vast supplies stored in London for the use of the Roman troops.

While Suetonius thought only of pushing his vengeance to the uttermost, the new procurator, Julius Classicianus, thought of a province ruined and sources of revenue dried up. He begged Nero to recall Suetonius and send out a more humane governor, to practise a policy of conciliation and repair, instead of aggravating the effects of what now could not be undone. The em-peror sent a commission of inquiry under his freedman secretary Polyclitus, with orders to settle the differences between Classi-cianus and Suetonius and to see to the pacification of the rebels; for in spite of the governor's ruthlessness, the tribes had even now made no formal submission, and it was said that they were holding out on the procurator's advice, in hopes of better terms from another governor.

The commission decided that enough punishment had been inflicted. Suetonius, to whose soldierly mind it seemed that clemency towards natives was only a name for weakness, refused to accept the decision; and Nero had to recall him, to his bitter resentment, on the pretext of a mishap to his fleet. The cam-paign against the rebels was stopped, and in the autumn there arrived a new governor, C. Petronius Turpilianus, fresh from his consulship in the first half of the year, with instructions to conciliate his subjects and keep the peace. These instructions he faithfully obeyed.

In the pages of Tacitus, whose incomparable prose has given us the story of these years, Suetonius is a hero despised and rejected by a wicked emperor, and Classicianus and Turpilianus are abject and self-seeking cowards. But the verdict of Tacitus cannot stand. In the time of Nero, Britain certainly still needed soldiers like Suetonius, but she also needed statesmen to guide and foster the peaceful life of her Romanized tribes: parcere subiectis as well as debellare superbos. The story of Caratacus is the story of a gallant losing fight, honourable alike to victors and vanquished. The story of Boudicca from first to last is a story

104

CARATACUS AND BOUDICCA

of horror and shame, a story of things that ought never to have

happened. For the people of Britain, Caratacus has always and

justly been a national hero. It may seem harsh to refuse the

same title to Boudicca. But the real hero of her story is Classi-

cianus, who stood up to Suetonius in his hour of victory as the

champion of the British people. Classicianus died before he laid

down his office, and his ashes were laid to rest in London. His

tomb1 is among the possessions of the English nation; and in

honouring its stones we are honouring the man by whose work

that story, opening in violence and wrong, was made to end in

the establishment of a lasting peace.

1 One part of it, found on Tower Hill in 1852, has long been in the British

Museum, where it has now been joined by another piece of the inscription, found

in 1935, which makes its identification certain. Cf. Antiquaries Journal, xvi (1936),

pp. 1, 208.

VII

FROM BOUDICCA TO AGRICOLA

F

OR ten years after Boudicca’s rebellion history tells us little

of British affairs. But its very silence is informative. Tacitus,

who dismisses these years in a few scornful sentences,1 neglects

them because they were a time of peace, when Roman arms

were gaining no glory in frontier-wars, and when Roman gover-

nors were disgracing their office by conciliating their subjects

and allowing their sword to rust unused. If warfare is but an

instrument of policy, and if the end of policy is the establishment

of peace and prosperity, the achievements of Petronius Turpi-

lianus and of Trebellius Maximus, who succeeded him in 63,

were more important than those of the soldier-governors whom

Tacitus praised. But of these achievements in detail we know

nothing. Hints, hardly amounting to proofs, are offered by

archaeology of peaceful development and growing civilization

in the British country-side at this time. The towns destroyed by

Boudicca were rebuilt and others enlarged. It was during this

decade that a cautious administration and financial economy

allowed the conquered Britons to settle down into contentment

and prosperity, to recover from the wounds that had both

excited and accompanied the Boudiccan rebellion, and to lay a

foundation for the more spectacular advance in romanization

which is associated with the name of Agricola.

The military situation had reached a point of equilibrium. In

the west the flames of the fire lit by Caratacus had died down.

Wales, still unconquered, was quiet, no longer threatening the

security of the lands east of the Severn. Brigantia was a client

kingdom. The Second legion at Gloucester, the Fourteenth and

Twentieth at Wroxeter, and the Ninth at Lincoln (the evidence

for placing them at these sites has already been considered) re-

mained stationary, and afforded with their attached auxiliaries

an ample garrison for the frontier-line.

If anything, the garrison was excessive. In 67, when Nero was

planning a great expedition to the East, he was so well satisfied

with the results of his peaceful policy in Britain that he judged

1

Agric. xvi, §§ 3–6. In the Histories he mentions Britain, with one exception

(iii. 45), only so far as concerns the intervention or non-intervention of its troops

in the civil wars.

106

FROM BOUDICCA TO AGRICOLA

three legions enough for its needs, and withdrew the Fourteenth,

which had won great fame by bearing the brunt of the campaign

against Boudicca, to be part of his expeditionary force.

But Nero's days were numbered. In the following spring

Julius Vindex, a romanized Gaul, set in motion the forces that

were to destroy him. Governor of one of the Gaulish provinces,

he raised a revolt against Nero and approached the army

commanders of the west for support. One of them, Sergius

Sulpicius Galba, Legate of Hispania Tarraconensis, was to be

the new emperor. It is hardly possible but that Vindex should

have appealed to the governor of Britain as well; and, if so, this

appeal may have been the occasion of the trouble, briefly re-

corded by Tacitus, that arose between Trebellius Maximus and

his legions. Trebellius was a man of no military experience, who

in his pursuit of peaceful administration had allowed his troops

to become idle and undisciplined. A mutiny broke out, in which

Trebellius at first barely escaped the swords of the legionaries,

and finally, giving way to their demands, resumed for a time

a nominal and ineffective control.1 Such an event might have

been caused by an intention on Trebellius's part of throwing in

his lot with Vindex, and a refusal by his legions to follow him.

In any case, during the later course of the civil wars it was often

said that the army of Britain would never appear on the Conti-

nent, as if its decision on that point had been once for all

expressed.

When Galba's short reign had begun and ended, and when

Otho was being attacked by Vitellius, the British legions could

no longer ignore the very existence of the civil war. The Four-

teenth, by now in Pannonia or Dalmatia, declared for Otho;

but the troops in Britain joined the party of Vitellius; and now

once more trouble broke out between the governor and his

legions. The ringleader against him was one Roscius Coelius,

legate of the Twentieth. Trebellius accused Roscius of in-

subordination and lax discipline; Roscius accused the governor,

whose financial policy made him unpopular with the men, of

impoverishing the troops. From the legions, disaffection spread

to the auxiliaries, and the governor was compelled to leave his

province and take refuge with Vitellius. Once more, it is a

1 Tacitus has two different stories: the above in Agric. xvi, § 5, another in Hist.

i. 60. In combining them I follow Anderson, Agric., loc. cit.; but perhaps the

Hist. version corrects the Agric.; cf. p. 101 note.

reasonable conjecture that Trebellius had wished to bring his

legions across the Channel to the support of the Vitellian armies,

and that their refusal was the cause of his flight. Though nomi-

nally on the side of Vitellius, not a man from the army of Britain

took part in the campaign which set him on the throne.

That campaign culminated in the battle of Bedriacum, near

Cremona, in April 69. The victory of the Vitellian armies was

complete. A mere vexillation of the Fourteenth had been pre-

sent to share the defeat of the Othonians, and the legion as a

whole, arriving in Italy soon afterwards, claimed to be still

unbeaten. Rather to get rid of it than because it was needed

there, Vitellius sent it back to Britain, sending out at the same

time a new governor, Vettius Bolanus. Since Trebellius’s flight

the province had been governed by the three legates of legions,

Roscius Coelius putting himself at their head. The legion went,

not without recalcitrance. Presumably it returned to its old

quarters. A later tombstone at Lincoln naming one of its veterans

is not evidence that in this, its second, residence in Britain it

shared that fortress with the Ninth.

Bolanus governed Britain in the same peaceful manner as his

two predecessors. But conditions there were beginning to change.

Cartimandua had divorced her husband Venutius and married

his armour-bearer Vellocatus, and, presuming on the high

favour she enjoyed with Rome since her betrayal of Caratacus,

had attempted to impose her new consort upon the tribe as

partner of her throne. But Venutius was a popular figure; now

Caratacus was gone he was regarded as the first warrior in

Britain; and the Brigantes were indignant at the queen’s treat-

ment of him and her preference for a squire. The tribal leaders

joined him in revolt against Cartimandua and repudiation of

the Roman alliance. For the third time Cartimandua appealed

to a Roman governor for help. Bolanus sent to her aid a mixed

force of auxiliaries, who after severe fighting rescued the queen

but left her realm in the hands of Venutius. Brigantia, after

being a precarious ally of Rome for twenty years and more, was

now an open enemy, and the military equilibrium of the British

frontier was at an end.

At the same moment storm-clouds in the east were threaten-

ing Vitellius. In the summer of 69 Titus Flavius Vespasianus,

governor of Judaea, who had commanded a legion in Britain

under Claudius, was declared emperor, and his partisans began

108

FROM BOUDICCA TO AGRICOLA

to march upon Italy, gathering legions as they went. Vitellius, collecting forces to resist them, took what he could from Britain. From the Fourteenth he had nothing to hope; it was actually in communication with Vespasian’s friends; but he had already some British troops with him, for when he marched into Italy after Bedriacum he was accompanied by 8,000 men from the Second, Ninth, and Twentieth.

He pressed Bolanus for auxiliaries as well, but the situation on the Brigantian marches was now such that none could be spared, and Bolanus answered evasively.

While the first of the Flavian armies was threatening Italy from the north-east, its commander, Antonius Primus, sent a message to the Batavian chief, Julius Civilis, requesting him to prevent the legions of the lower Rhine from coming to the help of Vitellius. Civilis promptly organized a rebellion which, beginning among his own tribesmen, soon had the whole Rhine-land aflame. Meanwhile Antonius entered Italy, met the Vitellians before Cremona, routed them, stormed their camp, and sacked the city.

By December Antonius was in Rome and Vitellius dead. One of the new emperor’s first tasks was to put down the rising of Civilis; in this work a leading part was taken by Q. Petillius Cerialis, a man both formerly and afterwards celebrated in British history, and by the Fourteenth legion, re-called from Britain for that express purpose and never again sent back there.

Vespasian’s rule was accepted in Britain without serious demur. His old corps, the Second legion, was eagerly in his favour and carried the others with it. For the moment, Bolanus was left in command. Though a Vitellian, he had shown himself a lukewarm one, more intent on governing his province than on fighting the private battles of the emperor who had appointed him. But Vespasian had no intention of leaving him there permanently.

The fall of Cartimandua had radically changed the British situation. The largest tribe in Britain, separated only by the Humber and Trent from Roman territory, had become hostile, and was being ruled by a tried and honoured warrior. In such conditions it was useless attempting to maintain the peace that had lasted for nearly ten years. The wisest course would be to reinforce the army of Britain, to replace Bolanus by a first-rate soldier, and to take the offensive against Brigantia at the earliest possible moment, namely, the moment when the revolt of Civilis should be at an end.

CERIALIS ATTACKS BRIGANTIA

109

That event furnished both the troops and the man. Vespasian himself had, during the civil war, raised a new legion, II Adiutrix, from the marines of the fleet at Ravenna, when it came over to his side; in the Batavian rising this legion fought under Cerialis against Civilis, and after the final battle lay for a time in camp at Neumagen. It was not needed for the permanent defence of the Rhine, and by the spring of 71 it was ready to be used elsewhere. As for a new governor, especially appointed for a war against the Brigantes, nobody could be more fit than Cerialis himself. As a former commandant of the legion at Lincoln he knew Britain, and especially the north; he had fought there with conspicuous gallantry in the Boudiccan war; lately he had shown initiative and courage in his work for Vespasian during the last weeks of Vitellius's reign; and finally he had crushed Civilis and brought a dangerous and expensive revolt to an end. As soon as he was at liberty, therefore, Vespasian sent both him and the Second Adiutrix to Britain.

For three years Petillius Cerialis governed Britain. During that time he pressed the war against the Brigantes with ruthless energy, and ended, says Tacitus, by either annexing or over-running a great part of their country. The words are vague; and other sources can add little. Excavations have shown that the carlicst Roman occupation at York goes back to about this time, and make it very probable that, at the beginning of his governorship, Cerialis either moved the Ninth legion bodily to York, lodging it there in a new permanent fortress, or else established there a camp of the kind that antiquaries call semi-permanent, which a large force might use as its advanced base during a campaign. Perhaps a semi-permanent camp for the Ninth legion was constructed at York in 71, followed by a permanent fortress a year or two later. In any case, we may ascribe to Cerialis the beginning of what was to be, until the close of the Roman age, the principal military centre of northern Britain.

The choice of York for this centre was a good one. Up to that point the Ouse is navigable to vessels of considerable size; and there the vale of York is crossed by a ridge of relatively high ground, safely above flood-level, dividing the low-lying forest and swamp to southward from other marshy ground to northward. This ridge is the true centre for all offensive strategy in north-eastern England. It is not a defensive position; it is not a position at which to block the movements of invading enemies;

110

FROM BOUDICCA TO AGRICOLA

it is a position from which to strike. Two natural lines of communication run north and south along the vale of York, both

used by Roman roads: one on the west, from Doncaster by Wetherby to Catterick, one on the east, from Brough-on-Humber

by Stamford Bridge to Northallerton. York lies midway between them, and is the only point from which both are easily

accessible. Road traffic between York and Lincoln, or between York and the north, can use either indifferently. Eastward all

the Wolds and Cleveland hills are within two days’ march; westward the Aire gap, the easiest of all the Pennine passes, gives

direct access to the Lancashire plain.

Another large camp of the same period has been identified and partly dug seventeen miles farther to the north-east, at Mal-

ton. This lies in the very centre of those eastern uplands which in prehistoric times were the richest part of Yorkshire;1 and a

general intending to cripple the Brigantes by ravaging their best agricultural land might very well have established himself here

even before he seized the strategic position on the Ouse.

Elsewhere no certain traces of Petillius’s campaigns have been found. The Brigantian power was of such a kind that an effec-

tive blow at its heart would paralyse its whole body. Its warriors once decisively beaten, its richest lands in central and eastern

Yorkshire occupied, and its fortress at Stanwick destroyed, further resistance was impossible. To a skilled and experienced

soldier, with forces enough at his disposal, this was no impossible task in three years’ fighting; and in this sense, even if he left

much of their outlying territory untouched, the conquest of Brigantia was complete by the time Cerialis laid down office.

Thus triumphantly begun, the forward movement went on. Ever since the time of Ostorius Scapula, whose frontier on its

northern wing had now been left far behind, the Brigantes and Silures had been the two chief dangers to the security of Roman

Britain. The Brigantes conquered, it was the turn of the Silures.

In 74 Cerialis was succeeded by another able and energetic governor, Sextus Julius Frontinus, a man greatly respected by

his contemporaries, and still represented by his writings on warfare, on surveying, and, in particular, on the water-supply of

Rome, a service which owed much to the vigorous reorganization and reforms carried out when he was at its head. Petillius, so

far as we can envisage him, was a pure soldier; Frontinus was

1 On this district, see Richmond, Arch. Aeliana, ser. 4, xxiii (1936), pp. 325-7.

FRONTINUS IN WALES

a man of high distinction as lawyer, engineer, and administrator; and we may be sure that during his governorship there was progress in the political and economic life of Britain as well as in the military affairs of which alone information has reached us.

Even that information is scanty enough: a single sentence, in which Tacitus writes that Frontinus ‘conquered the powerful and warlike Silures, overcoming both the valour of his enemies and the difficulty of the ground’; and a later passage in which he tells us that the Ordovices of central and north-western Wales had, before Agricola’s arrival, almost annihilated a cavalry regiment stationed in their territory, and that a single short campaign by Agricola sufficed to reduce them.

This implies that Frontinus, in addition to overpowering the Silures, undertook considerable operations farther afield, planted garrisons among the Ordovices, and by the end of his governorship left Wales as a whole conquered, though still able to rebel.

Archæology adds confirmation and a few details. Excavation in the legionary fortress at Caerleon-on-Usk has given ample proof that in its original form, with clay ramparts and wooden buildings, it was built about 75.

As Cerialis pushed his right-wing legion forward from Lincoln to York, planting it in the centre of the tribal territory he was attacking, so, it is clear, Frontinus moved up his left-wing legion from Gloucester to Caerleon in the heart of the Silurian country.

Here again the choice of a site was well made, and the new fortress became a permanent feature of Roman Britain.

The tidal Usk, like the Ouse, gave an easy approach for shipping. Alternative roads, either by the coast or inland through Monmouth, led in two days’ march to Gloucester.

Wroxeter lay not more than twice that distance by easy ways to northward. As an offensive centre, Caerleon lies on a coastal road leading into Pembrokeshire and commanding all the valleys of the southern Welsh hills, and in the vale of Usk, the best of all lines for penetrating the Silurian hinterland.

‘No other site in south-eastern Wales is comparable in respect of general accessibility from every point of the compass.’1

A few miles to the east, a hill-fort in Llanmelin Wood marks the probable site of the chief Silurian stronghold.

The settlement here, founded by Celtic colonists from across the Severn, apparently in the second century before Christ, was subsequently enlarged, and towards the middle of the first century A.D. its

1 R. E. M. Wheeler, Prehistoric and Roman Wales (1925), p. 223.

112

FROM BOUDICCA TO AGRICOLA

defences were improved. About the year 75 occupation ceases abruptly, and never begins again except for the building of a

few medieval huts in and about the site. At about the same time the first traces of occupation appear in the town of Caerwent on

the plain below. Frontinus, when he conquered the Silures, did not destroy the tribe as a political entity, but built them a new

capital where, under the eyes of the legion, they could learn to live like Romans and manage their own affairs in their own

city, henceforth called Venta1 of the Silures.

Westward and northward Frontinus built roads and studded them with forts. The best known of these, dated with fair cer-

tainty by excavation to his rule, is the Gaer near Brecon in the vale of Usk: ‘A pleasant, sheltered place where a permanent

garrison might live with tolerable safety and in ample comfort.’2 Another of his posts has been identified as far afield as Pem-

brokeshire. Other forts to south and north of Brecon, Coelbren in Brecknock and Castell Collen in Radnorshire, are probably

of the same time; and it is likely that further digging would give similar results at many other forts in South Wales.

To Frontinus, in view of what Tacitus tells us, we may also ascribe the early Flavian forts in the upper Severn valley, mark-

ing a line of penetration from Wroxeter into the country of the Ordovices: the Forden Gaer near Montgomery, and Caersws,

the strategic key to central Wales, where to-day the roads divide that lead left and right of Plynlimmon to Aberystwyth and

Machynlleth. Even in the extreme north, Caerhun and Segontium by Carnarvon are archaeologically dated either to Fronti-

nus or very soon afterwards: but here the evidence points, on the whole, rather to Agricola than to his predecessor.

There is a curious contrast between the abundance of archaeo-

logical evidence for the work of Frontinus in Wales and its paucity for that of Cerialis in Brigantia. To Cerialis we can

ascribe only the fortress at York and the great camp at Malton; not a single castellum dating from his campaigns has been identi-

fied; whereas Frontinus can claim not only the Caerleon fortress but, with tolerable certainty, half a dozen castella as well. Com-

parisons of this kind are dangerous; the difference may very well

1 Venta is a common Celtic town-name, but its meaning is unknown. It cannot be connected with the plausibly identical Spanish venta, an inn.

2 R. E. M. Wheeler, The Roman Fort near Brecon (Cymmrodorion Society, 1926), p. 69.

be due to the excellence of the work done by Welsh archaeologists at Roman military sites in the last ten years; but it is at least an interesting fact that, until the present time, only a single auxiliary fort (Margidunum) earlier than Frontinus has been discovered in Britain, although scores must have existed; and a possible reason might be that Frontinus, trained engineer that he was, introduced into Britain a better and more solidly constructed type of castellum, the type, in fact, of which Agricolan examples are well known and widely distributed. Of Agricola's castella, says Tacitus, not one was ever taken either by storm or by siege, so skilfully were they placed and so well provisioned; and it is conceivable that some part of his success in this matter may have been due to his adoption of improvements originally made by Frontinus.

It was late in the summer of 78, when Frontinus was succeeded by Gnaeus Julius Agricola, the third of Vespasian's fighting governors. His fame as a figure in British history rests in great part, doubtless, on his own merits; but in great part on the life which his son-in-law Tacitus wrote of him; so that it is difficult for the historian of to-day to keep a just proportion between his achievements and those of his predecessors, carent quia vate sacro. Nevertheless it must be attempted.

Agricola, born in the highly romanized Provence, had already spent much of his life in Britain. He had served there as a young man under Suetonius Paulinus, had fought through the Boudiccan rebellion, and had won a good name as a zealous officer, keen to learn all he could of the province and its army. He gained experience of administration in Asia Minor and Rome, and at the beginning of Vespasian's reign went back to Britain to command the Twentieth legion, suspected of disaffection. Under Bolanus, we are told, he learned to repress his thirst for military fame, but later commanded his legion in the Brigantian wars of Cerialis, after which he was sent to govern Aquitania and then, after a brief consulship, became governor of Britain. For all Tacitus's praise, he remains a somewhat unattractive character: an able man, evidently, both in the field and in administration; a man of sound judgement, and incorruptible; but cold, calculating, obsequious to authority, yet grudging in his submission; incapable of inspiring enthusiasm like Cerialis or reverence like Frontinus; in his last years, after a longer term of office in Britain than most, a man with a grievance, resenting

114

FROM BOUDICCA TO AGRICOLA

his recall, living in gloomy retirement for safety's sake, and

nursing his spleen.

His achievements in Britain were genuine and important; but

the rhetoric of Tacitus has tended to exaggerate them, and to

an incautious reader appears to give him credit for what in

reality was the work of his predecessors. It is only with an effort

that we recognize how completely Cerialis had already para-

lysed Brigantia and how nearly complete was Frontinus's con-

quest of Wales by the time Agricola reached Britain. His first

campaign, late in the summer of 78, finished the work of Fronti-

nus by conquering Anglesey and massacring the Ordovices; his

second, in 79, overran Brigantia and sprinkled it with a network

of forts. That is the year of the inscription on the leaden water-

pipes of the legionary fortress at Chester; and the construction

of the fortress itself should most probably be dated to the

beginning of Agricola's governorship, the Twentieth legion

and the Second Adiutrix being now placed there in prepara-

tion for his northward advance. His real forward movement

began in 80;1 in the next year he completed the conquest of

the Lowlands as far as the Forth and Clyde, studding the

country as he went with castella, built in timber and earthwork,

some meant for permanent police-forts, others designed to be

held only until the tide of conquest should have flowed on.

At the Forth-Clyde line he halted for a summer, partly for some

operation overseas which the text of Tacitus does not sufficiently

describe, partly to create a naval base (perhaps at Ravenglass

on the Cumberland coast) for a projected invasion of Ireland

which was never accomplished. In 83 he moved forward again,

marching into central Scotland by way of Stirling and encamp-

ing here and there upon navigable rivers in order to have his

army supported and supplied by his fleet. Tacitus ascribes this

movement to fear of attack from the Caledonian tribes, now

mustering in league to oppose him; but it is not to be thought

that Agricola would otherwise have been content to make the

Forth-Clyde line a permanent frontier. His ambition was cer-

tainly to conquer all Britain. What Tacitus means is that this

threatened hostile concentration induced him to cancel his plans

for an Irish campaign and proceed at once to take the offensive

in the north.

1 Novas gentes, Tacitus, Agric. xxii, § 1, implies that until then he had been going

over old ground.

AGRICOLA'S CALEDONIAN WAR

115

Agricola's plan for this summer was to carry a chain of posts

a certain distance along Strathmore, the great valley that separ-

ates the Perthshire mountains from the Ochil and Sidlaw hills:

then, resting this line on the sea (doubtless on the Tay near

Perth) to strike from it north-eastward up Strathmore itself and

north-westward up the valleys of the Earn and upper Tay. He

meant to establish himself in this offensive position before the

summer of 83 was at an end; and, hoping in this way to antici-

pate the enemy's concentration, he divided his forces into three

parts in order to push forward the work of occupying and fortify-

ing the entire position. In this he was successful, despite a

dangerous attack upon the Ninth legion, whose failure did

nothing to discourage the Caledonians. Next year their con-

centration was complete. The Mons Graupius, where it took

place, has never been successfully identified. If the above ac-

count of Agricola's doings in 83 is correct, the effect of that year's

campaign had been to force upon the Caledonians a place for

their concentration outside the area now occupied by the Roman

armies, that is, outside the region limited on the north by Inch-

tuthill, the Roman fortress at the confluence of the Tay and Isla.

Since the words of Tacitus1 suggest that it took place not very

far from the sea, it would therefore be natural to place the Mons

Graupius in or near the north-eastern part of Strathmore, near

Forfar or Brechin; and the importance of this region in early

Caledonian history is well enough attested by the existence,

close to Brechin, of two great hill-forts, the White Caterthun

and the Brown Caterthun, 'among the largest in Scotland,

placed on rival summits of an isolated ridge . . . with the

Highlands of Forfar at their back and the Lowlands of the

same county at their feet'.2

It was late in the summer when Agricola fought his famous

battle and destroyed the assembled armies of Caledonia. No

time was left to establish police-posts in the country of the de-

feated tribes: all he could do was to march, ostentatiously at

leisure, through as much of it as possible, while making his way

back to winter quarters, receiving hostages from some, and upon

others content to leave an impression of his might. Meanwhile

his fleet cruised northwards and ravaged the coast to its farthest

extremity before it, too, returned to port for the winter.

1 Praemissa classe, Agric. xxix, § 2.

2 Christison, Early Fortifications in Scotland (1898), p. 256.

116

FROM BOUDICCA TO AGRICOLA

In the following year Agricola was recalled. Tacitus alleges

that Domitian cut his term of office short from jealousy, fearing

a comparison between Agricola’s genuine success in Britain and

his own derisory claims to success in Germany, and thinking it

dangerous that a private citizen should thus outstrip an emperor

in military renown. Modern writers, estimating more justly the

importance of Domitian’s German campaigns and recognizing

that Agricola had already enjoyed more than the customary

length of office in Britain, have preferred to think that the em-

peror, having obtained independent advice about the prospects

of further conquest, judged it impossible for Agricola to carry

out his programme, and recalled him in order to prevent further

waste of the empire’s resources. This would explain Agricola’s

resentment, but it demands a very unnatural rendering of the

famous phrase in the Histories of Tacitus, perdomita Britannia et

statim omissa. When Tacitus writes that Britain was completely

conquered and then let go, he must on this view mean that

it was not conquered, but could have been conquered had

Agricola not been recalled.

Taken literally, the words of Tacitus imply that the battle of

Mons Graupius was a decisive victory, a final destruction of the

enemy’s armed forces, and the end of the war. There is plain

evidence that both he and Agricola so regarded it. Not only in

the Histories, but in the Agricola itself, the statement is explicitly

made that Agricola completed the conquest of Britain; and in

the governor’s own speech before the battle he is made to say

that this conquest would be finished by that day’s victory.1

Moreover, the opinion cannot have been peculiar to Agricola

and his friends. It is clear from precedents that, according to

the generally received theory of the time, a general who had

invaded an enemy’s country, found and destroyed his armed

forces, and planted permanent garrisons at strategic points there,

was regarded as having conquered it even without overrunning

and pacifying it in detail. To recognize this we need go back

no farther than Cerialis and Frontinus. Cerialis was recalled

after breaking the resistance of the Brigantes and establishing a

legion at York. There is no suggestion that he had a grievance

because he was not left in Britain until all Brigantia was overrun.

1 Tacitus, Agric. x, § 1: ‘tum primum perdomita est’; ibid. xxxiii, § 3: ‘finem

Britanniae non fama nec rumore, sed castris et armis tenemus: inventa Britannia

et subacta.’

PERDOMITA BRITANNIA

117

His position with regard to Brigantia in 74 was, in the eyes of contemporaries, exactly analogous to that of Agricola ten years later. Similarly, Frontinus at his recall had done all the decisive work towards the conquest of Wales; it is not suggested that Vespasian injured him by leaving the final settlement to Agricola. In short, Domitian treated Agricola strictly according to precedent in recalling him after his decisive victory, and not ungenerously in giving him seven years in which to achieve it. There is no reason to doubt the sincerity of the welcome which, according to Tacitus himself, he had ready for his returned servant.

But if Agricola, Domitian, Tacitus, and all other contemporaries thought it a decisive victory, it is even more difficult for modern historians to agree with their opinion than to realize that they sincerely held it. This is, no doubt, because Scotland has so often and so successfully resisted conquest that we find it hard to believe in the possibility of what, to these Romans, seemed easy. But the theatre of Scottish national resistance has always been the Lowlands. When they have been penetrated, when southern and central Scotland have been conquered, history does not teach us that the Highlands can resist by themselves. The nearest military parallel to the situation of Agricola after Mons Graupius is that of Cumberland after Culloden; and if Cumberland’s pacification of the Highlands was a grim business, that was because the political situation was complicated by partisanships and loyalties which had no existence in the time of Calgacus. If, as seems probable, the Caledonia of those days was ruled by a Celtic warrior-aristocracy like that of the Brigantes, whose peasantry had little kinship in culture and tradition with themselves, a single decisive battle entailing the destruction of that aristocracy would be the end of resistance.

Perdomita Britannia is, therefore, no rhetorical exaggeration. It expresses accurately the opinion of Romans at the time; and, so far as we can judge, that opinion was correct. Whether Agricola’s other opinion was correct, that Ireland could have been conquered with a single legion and a modest force of auxiliaries, is another question altogether. Agricola had never been to Ireland. But his success as a soldier entitles his opinion to be heard with more respect than it has commonly received.

Agricola’s real grievances against Domitian were two: that he

3720.I

I

118

FROM BOUDICCA TO AGRICOLA

was never employed again, and that Caledonia, which he had

conquered, was allowed to slip out of Rome’s grasp. On the

first head we can say nothing. As to the second, the main facts

are clear. They are, that Agricola’s military dispositions in

Scotland were designed by him as a base for further advance;

that this advance, if made forthwith, should, according to the

best judges, have been a simple matter, entailing no more pitched

battles; and that it was not made. It is equally clear that Domi-

tian’s reason for not making it was the concentration of his

military efforts elsewhere, on the Rhine and the Danube, efforts

which, about the year 86, entailed the removal of the Second

Adiutrix to Moesia and the permanent reduction of the British

army to three legions. By this change of policy Agricola’s wave

of Caledonian conquest was frozen into immobility the very

moment it had broken; and he ended his days feeling, truly

enough, that his crowning victory had been made fruitless.

The details are far less clear. Historians once believed that

immediately upon Agricola’s recall much of Scotland, if not the

whole of it, was evacuated. Early in the present century, exca-

vation at Newstead in the Tweed valley proved that here at

least there had been a garrison for some years after that date.

Following up this clue, Sir George Macdonald1 showed that the

same thing had happened farther north, at Camelon, Ardoch,

and Inchtuthill. He also showed that this continued occupation

had not been undisputed, but that these forts had from time to

time been roughly handled by native risings. How long this

state of things continued is not yet settled beyond dispute. In

his latest work he regards it as having probably outlasted the

reign of Domitian by at least eight or ten years, that is, until at

least 104–6; others, arguing from the absence of characteristic

Trajanic pottery, date its close before the end of the first cen-

tury.2 It would be more important than settling the exact date

when Agricola’s dispositions finally broke down if we could

know what exactly they were. The analogy of earlier conquests

compels us to believe that a legionary fortress, planted some-

where in Caledonian territory, was an integral part of his plan.

As long as there were four legions in Britain, this was possible;

1 ‘The Agricolan Occupation of North Britain’, in J.R.S. ix (1919). Cf. Macdonald on the Newstead evidence in Curle, A Roman Frontier Post (1911),

pp. 401, 415.

2 Macdonald, The Roman Wall in Scotland, ed. 2 (1934), p. 2; Pryce and Birley,

'The First Roman Occupation of Scotland', in J.R.S. xxv (1935).

THE FORTRESS OF INCHTUTHILL

119

but we do not know whether a legion was ever placed there,

either by Agricola or by his successor. If it was, it must have

been withdrawn at latest in 86 when the British legions were

reduced to three. At Inchtuthill, however, the 'great camp' of

fifty acres, with its twenty-foot ditch, its massive rampart, its

built roads, and its wooden hutments, forcibly recalls the per-

manent legionary hiberna of the Flavian period; and although

further excavation is needed to assess its meaning more accur-

ately, the most probable explanation of the discoveries made

there is that it was the fortress of a legion, built in the year 83

or 84 and evacuated either on Agricola's recall or, at latest, in

86.1 Even then the place was not deserted, for there are traces

of an auxiliary garrison living on the same site after the abandon-

ment of the great camp. But if this reading of the evidence is

correct, the withdrawal of a legion from Strathmore, with its

inevitable result in the fatal weakening of the whole position

there, was the virtual evacuation of Caledonia.2

1 The excavations (1901) are reported in Proc. Soc. Ant. Scot. xxxvi. Sir George

Macdonald, reconsidering their significance in his paper of 1919 already quoted,

concludes that the great camp was 'the winter quarters of a small army . . . in all

probability Agricola's' (J.R.S. xix. 115), and kindly informs the writer that his

meaning was, and is, to explain Inchtuthill as a permanent fortress.

2 In this chapter I have opted for the later of the two datings (77 or 78) for

the beginning of Agricola's term of office in Britain and consequently (since its

internal chronology is fixed) for all its incidents. I have been induced to do so

by Mr. Eric Birley's pointing out to me that by Agric. xxix the battle of Mons

Graupius was fought after Domitian's triumph over the Chatti, which must be

dated late in 83 (Camb. Anc. Hist. xi (1936), p. 164, note 2).

VIII

THE MAKING OF THE FRONTIER

Of British affairs for a generation after Agricola’s recall ancient writers tell us little or nothing. During the latter

part of Domitian’s reign a governor of Britain, Sallustius Lucullus, was put to death for daring to call a new type of lance by

his own name; apart from that scrap of scandal prescrved by Suetonius, we are left to read the story as best we can from

archaeological evidence until the accession of Hadrian.

Inscriptions give us the names of several governors. In 98, the second year of Trajan’s reign, a certain Nepos, successor

perhaps of the ill-fated Lucullus, had just laid down that office and had been followed by T. Avidius Quietus, whom we know

as a friend of the younger Pliny. By January 103 the governor of Britain was L. Neratius Marcellus, brother of the distinguished

jurist Neratius Priscus whom Trajan, according to Roman gossip, intended for his own successor on the Imperial throne.

Marcellus’s term of office was then probably nearing its end. The next name is that of Q. Pompeius Falco, who was presum-

ably the first governor appointed by Hadrian, and was immediately followed by A. Platorius Nepos, who supervised the

building of the Wall.

Another officer of some importance, about the same time, is a commander of the legion at Chester, T. Pomponius Mam-

milianus Rufus Antistianus Funisulanus Vettonianus. The inscription mentioning him used to be dated to the late second

century, but this is an error;1 and the officer named by it was perhaps a brother or relative of Pliny’s friend Mamilianus,

who was consul in 100, if not the same man.

Nothing can be said as to the doings of any one of Trajan’s governors. But excavation has told us something about military

developments during that reign, though it is not easy to put the details together into a coherent picture. The focus of interest

falls partly on Trajan’s Dacian and eastern wars, partly on his internal administration; and nothing happened in Britain,

1 As Haverfield saw: Ephemeris Epigraphica, ix. 535. The inscription (C.I.L. vii. 164) closely resembles in style others from northern Britain which date from

Hadrian’s reign: e.g. C.I.L. vii. 371, 373, set up by the Hadrianic garrison of Maryport.

THE MEANING OF LIMES

121

whether in peace or war, comparable with Trajan's great undertakings elsewhere. Conqueror though he was, he did not regard

the resumption of Agricola's northern conquests as an object worthy of his attention; on that adventure he accepted Domitian's verdict, and his military policy in Britain is best understood as an anticipation of Hadrian's: the policy of establishing a permanent frontier and renouncing all conquest beyond it.

The final result of this policy, as consummated by Hadrian, was to cut Great Britain in two by means of a continuous rampart and ditch, garrisoned by troops housed in forts along its line and provided also with fortlets at regular intervals, the so-called milecastles, and signal-towers between them. But before considering the steps by which this final result came about, it is necessary to understand the Roman conception of a frontier and to know something about the history of the Imperial frontiers in general.

To the Romans the continuous lines of fortification which are so much the most conspicuous feature of the British frontier were no necessary part of a frontier as such. Their word for a frontier, limes, originally meant a pathway; in military usage it means the strip of cleared and open land in which runs the via, the made and metalled road: and hence it comes to mean that strip with everything which it contains, road, forts, and signal-towers. Thus limes becomes the regular word for a military road complete with all these appurtenances, irrespectively of the direction in which it runs, whether forwards to the front of one's position, or transversely across it. This is the stage which the development of the word's meaning has reached by the earliest years of the empire, when military writers freely use it to signify a strategic road driven forwards into the enemy's country, like Agricola's road through Northumberland into central Scotland. But in the Flavian age, when frontiers began to be more thought of than conquests, it comes to bear a special sense, the sense of a frontier, conceived as a road of this same type running transversely and marking the limit of Roman occupation.

Throughout the history of the Roman empire, the Imperial frontiers for the most part retained this character. In Africa and in the east there are none of any other type. But often, where it was convenient, such frontier-roads were placed on the bank of a great river, which gave at once a convenient line of

122

THE MAKING OF THE FRONTIER

demarcation and a certain obstacle to unauthorized crossing

of the line. Augustus, who first took in hand the organization of

a frontier system for the Roman world, used the Rhine and the

Danube in this way; his attempted conquests in Germany were

designed to replace the Rhine by the Elbe.

During the German wars of the early first century, the Romans

encountered a different type of frontier. Some of the German

tribes surrounded their territory with continuous earthworks.

One of these, found in recent excavations, proves to have been

an earthen bank thirty feet wide and faced in front with a

massive upright revetment of timber. Similar tribal boundaries,

as we have seen in an earlier chapter, were introduced into

Britain by the Belgae in the first century before Christ. These

works appear to have given the Romans the idea of adding upon

occasion to their own standard type of frontier, the fortified

road, a new feature, namely, a continuous rampart. Such an

addition never became general. We have certain knowledge of

its existence only in three cases: from the Rhine to the Danube

in upper Germany and Raetia; across the Dobruja from the

Danube to the Black Sea; and in Britain.

Compound frontiers of this type, combining the idea of a road

with its forts and signal-towers and the idea of a continuous

obstacle, began to be constructed in the reign of Domitian. To

this reign, and probably to the years 87-9, belongs the great

earth rampart of the Dobruja, a work about 50 feet thick and

still standing in places as much as twelve feet high. It is fur-

nished with an extraordinary number of forts, surrounded by

turf ramparts, which have been occupied only for a short time

and then superseded by others smaller in size and fewer in

number; and alike in the massiveness of its construction and the

strength of its garrison it suggests an emergency measure fol-

lowing on a military disaster, which has been identified with

the defeat of Cornelius Fuscus in 87. The earthwork itself is

reasonably explained as an enlarged imitation of an earlier one,

following almost the same line, which is probably a tribal

boundary. The whole line must have been abandoned when,

at the beginning of the second century, Trajan planted a

legion sixty miles north of it, at Troesmis on the bank of the

Danube.

It was Domitian also who first erected a continuous frontier-

barrier in Germany; but this was of a much simpler type. His

FRONTIERS IN THE REIGN OF DOMITIAN

war against the Chatti in 83 resulted in the building of several

forts in the Main plain, at and around Frankfurt, and the enclos-

ing of this whole area with a frontier carried along the water-

shed of the Taunus and driven in a straight line across the

lower ground. This frontier was composed of a string of earth

fortlets connected by wooden signal-towers, and in front of these

a wattle fence. Later, all this was reconstructed after the rising

of Saturninus in 89, and at the same time the area enclosed was

extended both to north and to south.

Thus, by the end of the first century, the standard pattern

of Roman frontier was well established: a road, with forts and

signal-towers along it, which might run either across country or

along a river-bank.; and precedents were already established for

adding in front of this, as occasion seemed to demand, some

continuous obstacle, whether fence or earthwork. But with

regard to this addition there was no settled rule, either as to its

existence or as to its nature; it would not be added at all unless

local conditions required it, and if so its character would be

what seemed to suit the place and the time. Each provincial

governor, if called upon to construct a frontier, would use his

own judgement as to how it should be done, unhampered by

precedent except in the one essential, that it should be a road

with forts and signal-towers at intervals, though each governor

would doubtless know what had been done in other provinces.

At the end of the last chapter it was pointed out that the

removal of the Second Adiutrix from Britain to Moesia about 86

made it impossible for the British army to hold all Agricola's

conquests for long, unless the frontier tribes should remain

quiet. Evidence of destruction and rebuilding at sites in central

Scotland shows that they did not. Until about the close of the

century, however, it is certain that these conquests were main-

tained: at some time after that they were abandoned. There

is no reason to connect this abandonment with any further move-

ment of troops out of Britain, whether to Dacia or elsewhere;

the situation in northern Britain by itself is enough to explain it.

This abandonment of the Agricolan position in Scotland

could hardly be carried out at a time when so much thought

had already been given to the subject of frontier-lines without

the creation of a definite frontier somewhere farther to the

south: and this would presumably be a transverse road, with

forts and signal-towers, to which a continuous obstacle might or

124

THE MAKING OF THE FRONTIER

might not be added. For such a frontier either of two positions would be especially suitable: the Forth-Clyde isthmus, where Agricola himself had set up a temporary frontier-line in his campaign of the year 80, or the Tyne-Solway line, which was similarly used by him, it would seem, two years earlier, and where finally the Hadrianic frontier was placed. The Forth-Clyde line was rejected: if Strathmore could not be held, the Forth-Clyde isthmus was little better. Geographical conditions therefore led almost inevitably to the Tyne and Solway.1

On the Tyne-Solway line there are three things, each of which might be regarded as in some sense a limes. There is Hadrian’s Wall, with its forts, milecastles, and turrets, and the military way running close behind it. There is the earthwork which English antiquaries call the Vallum, following the same line a little way south of the Wall. And there is the Stanegate, a fortified road a little way farther south again. The construction of the Wall, whatever problems of detail it may still present, is dated beyond dispute to the earlier part of Hadrian’s reign. If, then, we are looking for traces of a Trajanic frontier on or near the same line, the Vallum and the Stanegate must be examined.

The Vallum is a broad and deep, flat-bottomed ditch, whose upcast earth has been neatly arranged in two parallel mounds set back some twenty feet from its north and south sides. This symmetrical section makes it tactically neutral: no force of troops on one side of it has any advantage over one on the other. In spite of repeated and careful search, no trace of palisades belonging to it, on the mounds or elsewhere, has ever been found. Although, with its steep sides, it was in its original shape a formidable obstacle, it cannot be regarded as in any sense a work of defence or fortification. That it is of Roman date, and belongs, like the Wall, to the early second century, is established beyond doubt. It runs from the north bank of the Tyne, close to Newcastle (the exact termination is not known, but it seems to have been less than a mile upstream from the Roman bridge and fort there) to the south bank of the Solway at Bowness, a distance of about seventy English miles.

1 A suggestion was at one time made by Haverfield (Edinburgh Review, April 1911) that the frontier, after central Scotland was abandoned, was placed on the Tweed. It is, of course, possible that Newstead was held for a time after sites farther north were evacuated; but there is no evidence that anything like a limes was ever constructed in the Tweed valley.

TRAJAN'S FRONTIER AND THE VALLUM

125

Antiquaries of the last century tried to explain the Vallum as a southward defence for a strip of frontier-territory, varying in width but nowhere more than half a mile across, whose northern edge was the Wall. When a closer study of its original shape made this view impossible, Haverfield in the last years of the century proposed to regard it as a civil or legal boundary somehow corresponding to the Wall as a military obstacle. The difficulty of understanding why two continuous works were needed for these two purposes led later archaeologists to suggest that Hadrian's original frontier had been a line of forts and signal-towers connected by the Vallum, which they explained as a mere line of demarcation without any military purpose; and that to this original Hadrianic frontier-system the Wall had then been added as a military obstacle, in order to ease the work of the garrison.

Apart from various other difficulties, this suggestion was based on the assumption that the Vallum was older than the Wall; and of this no definite evidence had ever been found. The prima-facie evidence was, in fact, against it; for on several occasions, when the Vallum approaches a fort belonging to the Wall, it swerves from its usual straight line and makes a deviation southwards so as to avoid the fort. This looks as if the Vallum were either contemporary with the Wall, or later than the Wall; if it is to be regarded as earlier, the prima-facie evidence must be explained away, either by supposing that the forts standing in these deviations are earlier than the Wall and belong to the presumed Vallum stage of the frontier, or by supposing that they belong to the presumed Wall stage but occupy the sites of earlier and perhaps smaller forts contemporary with the Vallum.

Expedients like these made it theoretically possible to regard the Vallum as earlier than the Wall; and the question arises whether in that case the interval between the two might be sufficient for the Vallum to be the required Trajanic frontier.1 But the balance of evidence is against this alternative. The forts that stand in the Vallum deviations are now known to date, in some cases and therefore probably in all, from Hadrian's reign, and no evidence exists that any of them superseded a Trajanic fort occupying the deviation before it.2 The Vallum itself has

1 This identification was proposed by Birley, ‘A note on the date of the Vallum’, in Arch. Aeliana, ser. 4, xi (1934), p. 146.

2 A fragment of what might be a small Vallum fort has been found under the

126

THE MAKING OF THE FRONTIER

in many places been purposely filled up to facilitate operations connected with the Wall; and botanists who have examined the material from the ditch at these places say that it has lain open only for a very short time, certainly not long enough for the interval between a Trajanic frontier and one dating from about 122. Lastly, recent work in Cumberland has excluded the possibility that the Vallum can be at all, however little, earlier than the Wall: for it is now known to make a deviation in order to avoid, not a fort, but a milecastle belonging to the Hadrianic series, built on ground where no earlier structure ever stood. It is therefore necessary to look for the Trajanic limes elsewhere than in the Vallum; and this means asking whether it can be more plausibly identified with the Stanegate.

The Stanegate is a road which runs from Carlisle along the north bank of the Eden by Low and High Crosby to Irthington, where it crosses the Irthing and passes close by a fort at Brampton Old Church; passing another fort at Easy, it travels up the left bank of the Irthing to another at Nether Denton and another at Throp, close to Gilsland, whence it climbs to the heights north of the Tyne valley to forts at Carvoran, Haltwhistle Burn, Chesterholm, and Newbrough. After this its course is uncertain. It may cross the North Tyne near Chesters; it may continue down the valley to Corbridge.

On this line decorated ‘Samian’ pottery of Trajan’s reign is common. Considerable quantities of it have been found at Carlisle, Nether Denton, Birdoswald (whose oldest structural remains, earlier than the Vallum, would seem to indicate an outpost of the same system north of the Irthing gorge), Chesterholm, Chesters, and Corbridge. Rougher pottery, dated to the same period, has been found at Throp and Haltwhistle Burn. Thus all along the Stanegate evidence of occupation in Trajan’s reign is definite and voluminous. That road itself, in some form or other, apparently goes back as far as Agricola; some of the sites on its line, notably Carlisle, Chesterholm, and Corbridge (if Corbridge may be regarded as belonging to it), undoubtedly

stone fort at Birdoswald (Cumb. and West. Trans., N.S., xxxiii (1933), pp. 253-4), but no evidence was forthcoming for dating it to Trajan; pottery of that reign was found close by (ibid. xxii (1932), p. 141), but this might equally well have come from the buildings that stood on the same site before the Vallum was made (ibid. xxxiv (1934), pp. 120–4). Nor is there any reason for ascribing a Trajanic date to the scrap of turf rampart at Castlesteads (ibid. xxxiv, pp. 163–4). The evidence for Trajanic Vallum forts, at the time of writing this, is nil.

TRAJAN AND THE STANEGATE

127

do so; but the systematic fortification of it belongs without a

doubt to the time of Trajan. The Stanegate, then, would appear

to be the limes constructed in that reign after the evacuation of

Scotland. In the absence of literary records and inscriptions no

closer date can be given; and we do not yet know how this limes

ran east of the North Tyne. It may possibly have followed the

road known as the Devil's Causeway to near Berwick-on-Tweed;

but until the course of that road has been explored with the

spade this must remain mere guesswork.

Elsewhere in Britain there are traces of a widespread reorgani-

zation in Trajan's reign. To some extent this was a matter of

routine. Hitherto, so far as we know, all Roman military build-

ings in Britain had been of timber, with earth or turf ramparts;

stonework seems hardly to have been used except in bath-houses,

where the hypocaust system of heating required it. But at least

by Domitian's reign earthen forts were being replaced in Ger-

many by stone; and in Trajan's time this fashion spread to

Britain. It was now that the legionary fortresses at Caerleon

and York, and probably at Chester, too, received stone-revetted

ramparts and stone internal buildings. At York, an inscription

enables us to date the change, or at least some important phase

of it, to 108-9; at Caerleon, another tells us that it was in pro-

gress about the turn of the century. Many auxiliary forts were

treated in the same manner. The best examples are in Wales,

where Segontium, near Carnarvon, received stone internal build-

ings, Caerhun in the Conway valley both these and stone ram-

parts, and Brecon stone ramparts and a beginning of stone

internal buildings, never completed.

More important is evidence suggesting that garrisons were

being moved about. At Castleshaw, in the hills east of Man-

chester, a large fort was abandoned at about the beginning of

the century, and a very small blockhouse built on its site. Cer-

tain reductions in the size of forts, made by cutting off one end,

have been tentatively ascribed to this reign, but without definite

proof: examples are Castell Collen and Tomen-y-Mur in Wales.

This need not mean that Trajan was withdrawing auxiliary units

from Britain for his wars in the east, for at the same time we

find certain new forts being erected, such as Hardknot Castle

in Cumberland and Gellygaer in Glamorgan, the latter dated

by an inscription to between 105 and 112. Such as it is, all this

suggests that, over and above a tendency to reconstruct old forts

128

THE MAKING OF THE FRONTIER

partially or wholly in stone (perhaps in some cases after enemy destruction, but more probably after natural decay of the wooden buildings), the reign of Trajan witnessed two shufflings of auxiliary regiments, which may or may not have been simultaneous: a movement of certain units from their Agricola

quarters in the far north to the new Stanegate limes, and, farther south, a shifting of others from forts which it was now decided to evacuate into others now built for the first time.

On the whole it is probable that these movements and changes were due rather to a spirit of reorganization, characteristic of Trajan, than to necessity imposed by military disasters on a large scale. Had there been a great rising in Britain during this reign it is probable that we should have heard of it from some ancient writer. Small local disturbances there may well have been. Now and then, an auxiliary fort may have been destroyed. But the only reason for imagining a disaster of any magnitude is the unexplained disappearance of the Ninth legion. The last dated record of its existence is the inscription of 108-9, already mentioned, at York. By about 122, when Hadrian's Wall was being built, it was replaced at York by another legion, the Sixth Victrix from Vetera on the lower Rhine, and had ceased to exist as an effective corps. But its disappearance cannot be dated earlier than the beginning of Hadrian's reign, when the Stanegate frontier had been for some time in existence; its annihilation cannot therefore have been any part of the events leading up to the evacuation of Scotland and the establishment of that frontier.

When Trajan died in the summer of 117 on his way back to Rome from the east, he was succeeded by P. Aelius Hadrianus, a native like himself of Italica, near Seville, his cousin by birth, his kinsman by marriage, his ward in boyhood, and his life-long friend. As a young officer in the legions of the Rhine and Danube, fighting with Trajan through the Dacian and Parthian wars, first on his staff and later in command of a legion, and then governing lower Pannonia, Hadrian had served long in a sound military school and had seen almost every part of the European and eastern frontiers. Promoted somewhat rapidly by the emperor's favour, he was at Trajan's death governor of Syria, the most important military command in the empire. He was not, like his predecessor, a soldier of genius; but a good soldier he was, and much more besides : to the solid traditions of soldier-

ship and statesmanship which he inherited from men like Trajan

and Vespasian he added something comparable with the scholarship of a Claudius and the artistic interests of a Nero, but mixed

these elements in such a way as never to incur the charge of a

pedantic learning or a frivolous aestheticism. He not only

created a stable and efficient system of imperial frontiers, he

also inaugurated a new age in the history of Roman taste; he

was both a patron of original ideas in architecture and himself

an architect of originality; he was a painter and a lover of

painting; a lover of literature and a writer who, after a lifetime

of idle versification, wrote on his deathbed five lines of poetry.

Except for one brief visit to the Danube, Hadrian spent the

early years of his reign in the capital. Then, in 120 or 121, he

set out on the first of the voyages that were to occupy the next

ten years of his life. The account of these voyages that has come

down to us is fragmentary and confused, but it gives us a clear

enough picture of a man tirelessly energetic, bewilderingly ver-

satile, throwing himself with equal energy into the adornment

of towns, the redressing of peasants’ grievances, the reviewing of

troops, or the inspection, afoot and bare-headed in all weathers,

of legion upon legion of frontier. Wherever he went Hadrian

left his mark. New laws were enacted, new buildings rose from

the ground. But nowhere has he lcft a more impressive monu-

ment than in Britain.

During the first few years of his reign matters there had not

gone too well. At the time of his accession, says one ancient

writer, the Britons were in successful rebellion against Roman

rule; and another says that under Hadrian the Romans suffered

heavy casualties in Britain.1 The two statements must refer to

one and the same event, namely, a rising in Britain, accom-

panied by a grave disaster to the Roman arms, in the first two

or three years of the reign. With that disaster must be connected

the disappearance of the Ninth legion.2 Consequently the

1 Scriptores Hist. Augustae, Spartian, Hadr. § 5; Fronto, p. 218 N.

2 Ritterling, in Pauly-Wissowa, Realencyclopädie, s.v. ‘Legio’, cols. 1668-9, has

pointed out that this connexion, although he does not reject it, gives rise to a

difficulty: for certain persons are known to us who served as tribunes in the Ninth

and thereafter passed through other stages in the senatorial career at a rate which,

if that legion really ceased to exist in 117–19, was extraordinarily slow. He asks,

therefore, whether the annihilation of the legion might not have happened in a

revolt at some later date in Hadrian’s reign. To my mind, its absence from the

inscriptions of the Wall, plus the fact that Hadrian brought the Sixth to Britain

(surely to replace it), makes that impossible. In any case, the existence of these

130

THE MAKING OF THE FRONTIER

situation with which Hadrian had to deal in Britain was not unlike that which had confronted Domitian in 87 in the Dobruja:

a badly defeated frontier-army to be reinforced, its morale restored, and its positions secured against further attack. This serves to explain why Hadrian’s British frontier works are in point of strength so much like Domitian’s in the Dobruja, and so unlike Hadrian’s own works in Germany.

It was from Germany that Hadrian’s British journey began; and it is instructive to see how he reorganized the Domitianic limes there before ever he visited Britain. The whole frontier-line was surveyed afresh and in some places moved to a better position. The auxiliary regiments acting as its garrison, instead of lying in forts at some distance behind the line, were brought up to fresh ones built upon the line itself. Finally, the entire line was defended by means of an oaken palisade, a massive work, still to be traced by the stumps of its uprights below ground-level, but rather an obstacle to unauthorized traffic than a fortification proper designed to stop the inroads of hostile armies. The aims governing these alterations are greater uniformity, greater efficiency, greater ease of working : and the chief means by which these results were achieved were the erection of the palisade and the placing of its garrisons as close to it as possible.

In certain ways Hadrian’s reorganization of the British frontier resembled these changes in the German. Obviously he surveyed the whole line afresh, and, if a friend’s reference1 to his ‘British walking-tour’ is to be taken in its natural sense, he did it not only in person but on foot. As we do not know how Trajan’s frontier ran on the east, we cannot tell how he altered it there, except that the bridge and fort on the Tyne at New-castle, being called Pons Aelius after him, are certainly his work; but in the centre he moved it up from the north slope of the

survivors is proof that the disappearance of the legion was not due to literal annihilation in the field. But a legion that had only suffered severe losses in action was normally brought up to strength by drafts from elsewhere; the suspicion arises, therefore, that the Ninth had not only been cut up, but had disgraced itself, which would account for its survivors’ slow promotion.

1 The friend was a certain Florus, who wrote in mock commiseration of the emperor’s hard life a doggerel running :

Ego nolo Caesar esse,

Ambulare per Britannos,

Scythicas pati pruinas.

GERMAN AND BRITISH FRONTIERS

191

Tyne valley to the heights of the Whin Sill, which even in

Trajan's time the troops must have actually patrolled; and on

the Irthing he transferred it bodily from the south side of the

valley to the north, where one or two outposts were already

being held. In both these cases he was following the principle

which he had observed in Germany when he moved the garri-

sons up from such low-lying forts as Wiesbaden and Frankfurt

to the ridge of the Taunus, increasing their efficiency even to

the detriment of their comfort. On the west he carried the line

down the south shore of the Solway to Bowness, where the firth

ceases to be fordable.

As in Germany, again, he built along this line a continuous

barrier. The country through which it was to run was not rich

in heavy timber, and therefore the barrier could not be a

palisade; but throughout its eastern and central part there was

abundance of good building-stone and of limestone for burning,

and in the western part, where stone became scarcer and no

lime was to be had, it was easy to build, as local conditions

prompted, either in cut sods or in clay. This was the original

plan for Hadrian's Wall in Britain. Beginning at the new bridge

and fort at Newcastle, the Wall, 10 Roman feet thick and 20

feet high, including its parapet, built of stone with a rubble and

mortar core, and protected in front by a 30-foot ditch, was to

run westward in this shape as far as the Irthing, 40 English

miles; then, with the change in geological conditions, the build

of the Wall was to change, and it was to go forward as a turf-

work, or made of clay where turf was not ready to hand, for

another 30 miles to Bowness-on-Solway. At regular intervals

of a Roman mile there were to be fortlets, walled in stone or

turf as the case might be; between each of these and the next,

two stone towers for signalling. These milecastles and turrets

(to give them their current names) were to be solid with the

Wall and of one build with it. At irregular intervals were six-

teen forts, almost all of them built actually in contact with the

Wall, and each housing an auxiliary regiment.

The chief difference between this Wall and the German pali-

sade is one of material. But there is also a very considerable

difference of strength. The British work is much the more

powerful, and, especially with its great ditch in front and its

rampart-walk on the top, has the air of a genuine fortification

intended to be used in warfare. Closer study shows that this

132

THE MAKING OF THE FRONTIER

appearance is deceptive. A wall-top 7 or 8 feet broad, to which

access was provided only by ladders every 500 yards and narrow

stairways a mile apart, lacking artillery, lacking bastions, and,

above all, garrisoned only with an average of 150 men to the

running mile, was never intended by Roman engineers for use

as a fighting-platform. Like the German palisade, it was meant

as an obstacle to raiding and plundering, and doubtless also to

smuggling; and since, if it were properly patrolled, no one

could cross it without attracting attention, especially when

laden with loot or merchandise, its efficiency must have been

perfect. The fact that it was so much stronger than the German

palisade no doubt reflects the fact, already noted, that the

British frontier was restless. The border tribes were in a danger-

ous mood; the Roman armies had lately suffered a grave defeat;

and in this respect conditions here resembled those which led

to the construction of Domitian's earth wall in the Dobruja,

thirty-five years earlier. Even granted that the Wall was in-

capable of resisting siege-tactics, it would be an obstacle to the

movements of armed bands, and would thus protect the Roman

posts against being surrounded, cut off, and crushed in detail,

while giving their garrisons a better chance of reinforcing

threatened points and marching out through the northern gates

of their forts to engage the enemy in the open.

As often happened with Roman buildings of great size, the

original plan of the Wall underwent modification before the

work was finished. In the east and centre it was decided to

reduce the thickness from 10 to 8 Roman feet, and at the same

time to extend the length for 3½ miles down the Tyne to a new

terminal fort, still called Wallsend, where a clear view can be

had down the last reach of the tidal river. When this change

was made, the stone Wall had been built westward from New-

castle for half its intended distance, as far as the north Tyne;

on the remainder of its line the foundations had been laid and

the milecastles and turrets built. Between the north Tyne and

the Irthing, therefore, the curtain of the Wall was completed

to the new 8-foot gauge. The decision was also made to re-

place the turf Wall by stone; this replacement, at the eastern

end of the turf Wall line, was put in hand at once, but it was

not completed for many years.

The Hadrianic frontier presents one other feature, the ditch

known as the Vallum, of which some account has already been

given.1 The purpose of this work has always been a perplexing

question; but close study of its actual features, partly as visible

on the surface of the ground and partly as revealed by the spade,

has gradually eliminated one hypothesis after another, narrow-

ing down the field of possibilities, until to-day the very stringency

of the conditions which a theory of its purpose must satisfy should

make it easier to see in what direction the truth is to be sought.

The chief facts to be considered are, that the Vallum in its

original shape was a formidable obstacle to traffic, but incapable

of military defence, and so designed, indeed, as to look ostenta-

tiously unmilitary; that this obstacle is carried with remarkable

thoroughness, admitting no interruption whether from hard

rock subsoil, morass, or ravine, right across the country from

Tyne to Solway, close behind the Wall; that, according to the

latest results of excavation, it was made at the same time as the

Wall itself; and that the only original ways across it are solid

causeways opposite the Wall forts and perhaps also opposite the

milecastles, each surmounted by a stone gateway. In sum, the

Vallum is a second obstacle parallel to the Wall and provided

with a corresponding series of controlled openings for traffic,

differing from it in its deliberately unmilitary design.

Now, a Roman frontier had two functions, one military or

defensive, the other financial, as a line where traffic passing be-

tween the province and the unconquered country outside it

passed through supervised openings and paid duty. And it is

a peculiar feature of Roman administration that the financial

service under the procurators was entirely separated from the

military service under the provincial governors. The sentry on

guard at the gate of a fort was responsible to the commandant;

he, whether or not through the mediation of a legionary com-

mander, to the governor of the province; and he to the emperor.

A customs officer was responsible to the procurator, and he to

the emperor directly, without any intervention of the governor.

And not only were these two services separate, but relations

between them were delicate: friction and jealousy were not

unknown.

Before the building of Hadrian’s Wall, continuous frontier-

works, where they existed at all, had been structurally separate

1 Above, p. 124. The name Vallum, as a piece of Latinity, is incorrect: Romans

would have called it fossa. But the current name has been applied to it ever since

the time of the Venerable Bede, and has earned a prescriptive right to stand.

3720.1

K

134

THE MAKING OF THE FRONTIER

from the forts of their garrisons. It may be conjectured (we have no proof) that the openings in the barrier were controlled by customs officers, while obviously the sentries of the garrisons looked after their own fort gateways. But on Hadrian’s Wall the forts, with one or two exceptions, formed part of the barrier itself, so that a man passing through the fort was passing the line of the barrier; and (again with one or two exceptions) there was no way of crossing the barrier except by thus going through a fort, unless, indeed, non-military traffic was allowed to pass through milecastles; if it was, the same problem would arise there.

From a military point of view, this new method of planning the forts in relation to the barrier was no doubt an improvement. If traffic crossing the line of the barrier was compelled to pass through a fort, the military control over such traffic was tightened. But the question must now have arisen, how to provide for the customs officers? Hadrian, a stickler for military discipline, may very well have thought it unwise to give the procurator’s men an official position at fort gateways, where the authority of the commandant should be undisputed. The simplest solution on paper, though a cumbrous and expensive one, would be to have a second barrier behind the Wall; to make this barrier look as unmilitary as possible, consistent with efficiency; and to provide it with a crossing opposite each fort, where the customs officers could do their work. The Wall as a whole would be controlled by the governor, the Vallum by the procurator; the distinction between the two reflecting and symbolizing the separation between the military and financial services.

There is no proof that this explanation of the Vallum is correct. All that can be claimed for it is that it fits the facts. The one thing which is certain is that, whatever may have been the original purpose of the Vallum, it was soon disused. Everywhere in the neighbourhood of forts the whole work was destroyed and its ditch filled in very soon after it was made; elsewhere, earthen causeways were thrown across it at short and regular intervals, which in some places were cleaned out again and in others left where they were. To the garrison, the whole thing was a nuisance; and it was no doubt discovered before long that the customs officers could, after all, be accommodated at the fort gateways without disastrous consequences; or, per-

THE BUILDING OF HADRIAN'S WALL

135

haps, in some cases, that even when the Vallum ditch had been

filled in the old gateway could still be used as a customs-house.

1

Hadrian's visit to Britain fell in 121 or 122. He left the pro-

vince in charge of A. Platorius Nepos, an intimate friend of his

own, who, since his consulship in 119, had been governing Lower

Germany; and it was this governor who carried out the con-

struction of the new frontier-works. The task was a heavy one.

To dig the ditch of the Wall alone, it was necessary to move

nearly two million cubic yards of soil and subsoil, much of it

solid rock; and in some places, where the rock is basalt, it can

still be seen how fiercely the ditch-diggers grappled with their

task, splitting the stone and lifting the blocks with cranes, until

they gave it up not as impossible but as an addition not, at such

cost, indispensable to the barrier of the Wall. The Wall itself,

with its core of rough stone and mortar—essentially a concrete

structure—faced on either side with ashlar, contained over two

million cubic yards of material, of which the stone had to be

quarried and cut, and the lime burnt. In Northumberland these

materials were to be had close at hand, and on certain stretches

of the Wall the Roman quarries are still visible not far behind

its line; in Cumberland, when the change was made from turf

to stone, a change which was probably begun under Platorius

himself, they had in many cases to be brought up from quarries

several miles to southward, where beds of suitable stone were

found exposed in steep river-banks. In addition, there were

rivers to be bridged, and about sixteen new forts to be built, all

in the most elaborate manner, of stone throughout. The highest

quality in construction was everywhere demanded; no trouble

or expense was spared to make details like fort-gateways and

bridge-abutments monumental alike in strength and magnifi-

cence; and beside all this, the Vallum has been calculated to

entail a million man-days of labour in mere earthwork. Even

to-day, when these vast works remain only in shattered frag-

ments, they are a monument of Roman purposiveness than

which none more impressive exists in any country, and a fitting

memorial to an emperor distinguished above others for his

ambition as an architect and his remorseless demands on the

labour of his troops.

1 The gateway long survived at Benwell (Arch. Aeliana, ser. 4, xi (1934), p. 182); per-

haps at Housesteads (ibid., p. 188) and certainly at Birdoswald (Cumb. and West. Trans.,

N.S., xxxiii (1933), p. 252) it was very soon disused. Practice was not rigidly uniform-

136

THE MAKING OF THE FRONTIER

For it was the legions that built the Wall. According to Roman tradition, every legion contained within itself technical specialists of every kind, and was competent to do every sort of constructional work that an army could need, whether temporary or permanent. Its efficiency as an engineering unit, for earthwork, woodwork, or masonry, was no less important than its efficiency in battle, and Hadrian himself, in his addresses to his troops in Africa, which have been preserved on stone where they were delivered, laid equal stress upon each. The sculptures on Trajan's Column, which may almost be regarded as contemporary pictures of the building of the Wall, show half-armed legionaries digging ditches and erecting fortifications, while auxiliaries act as covering troops. The Wall itself has yielded evidence of how this work was organized. Along its line over 150 small inscriptions have been found, generally cut on the ordinary stones of the facing, which simply give the name sometimes of a legion, sometimes of a cohort, but most often of a century, the smallest unit into which the legion was divided. They show that the work of building the Wall was partitioned out in lengths to individual centuries, just as, according to inscriptions from elsewhere, the ramparts of forts were divided. These lengths were quite small; now and then the inscriptions give measurements, which vary from 19 to 30½ passus (31 to 50 yards) but are most often between 35 and 40 yards; and the way in which the same centuries and legions reappear in different parts of the Wall proves that the whole, or at any rate the main, strength of the three legions was concentrated on building one short sector at a time, and then moved on to another.

Here and there, in the western part of the Wall, there are records of a different kind: once, near Birdoswald, a sector was built by men from the fleet; and in four places the work is recorded to have been done by British tribes: Durotriges, Dumnonii, Catuvellauni, and perhaps Brigantes, acting not as mere labour-gangs under legionary supervision, but as independent units signing their own work. But whereas the naval inscriptions are Hadrianic in style and must refer to the original building of the Wall (or rather, to the replacement of the turf Wall by that of stone), these tribal inscriptions seem all to be of later date, and refer to the rebuilding of the Wall in the time of Severus.

The completed work employed a large garrison. Each fort

THE GARRISON OF HADRIAN'S WALL

137

housed an auxiliary regiment, either of foot or of horse; the

total strength of these must have been, on paper, about 10,000

men, of whom probably one-fifth were cavalry. If we add the

forts on the Cumberland coast, which were certainly an integral

element in the Wall scheme, the depot at South Shields, and

the outposts north of the line, we get a total of about 14,000

men for the fighting garrison of the frontier. But the existence

of the milecastles and turrets, forming a system quite distinct

from the series of forts, suggests that there was, independently

of this fighting garrison, a patrolling garrison, one small unit of

which occupied each milecastle and served the two turrets im-

mediately adjacent to it. For the entire Wall, and the system of

signal-stations which prolonged its line down the Cumberland

coast, this patrolling garrison can hardly have numbered much

less than 5,000 men.

In order to understand the military situation thus created, it

is necessary to consider what troops a governor of Britain in the

early second century had at his disposal. We have no statistics;

but it is possible to draw up an army-list for Britain in Hadrian's

reign whose margin of error is probably not very large. This is

partly because inscriptions of the early second century are com-

mon; but it is even more because we happen to possess half a

dozen military diplomas dating from that half-century. These

were documents, engraved on bronze, issued to auxiliary soldiers

on their discharge, and certifying their good conduct while with

the colours and the conferment upon them of the rights of Roman

citizenship, which during their service auxiliaries, unlike legion-

aries, did not enjoy. In particular, the right of legal marriage

was always specified, though in such terms as to make it clear

that previous marriages, though not hitherto legally binding,

had already been recognized, and were now further dignified

with a legal sanction. These diplomas were issued in batches

to men of certain regiments; any one diploma, therefore, gives

a list of all the regiments to whose time-expired men copies of

it were issued; and when we put together seven lists of this kind,

all belonging to Britain, and dated respectively to 98, 103,

105, 122, 124, 135, and 146, we have a working basis for a

catalogue of auxiliary units in Britain for that time; to which,

in point of fact, only very small additions can be made from

other sources.

The results are as follows. Three legions, each numbering

138

THE MAKING OF THE FRONTIER

5,600 men, of whom 120 were cavalry, give 16,800 troops of

the first class. Of the 68 auxiliary units which we can identify,

16 are alae or cavalry regiments; one of these a milliary unit,

1,000 strong, the rest quingenary (nominally 500, actually on

paper 480): total 8,500. Of the auxiliary cohorts (infantry

regiments, but sometimes having a cavalry contingent) 7 are

milliary, 45 quingenary: total, 29,500. The auxiliaries thus

number 38,000 in all, and the grand total for the army in Britain

comes to 54,800. This, of course, is merely paper strength; we

do not know what percentage of it could be reckoned upon as

effectives. On the other hand, there may have been a few

auxiliary units of which we have not heard.

Of these 68 auxiliary regiments, the forts of the Wall itself

took up 16 or 17; and if we add the forts of the Cumberland

coast, the outposts in front, and the supporting posts within two

or three days' march of the frontier—a doubtful quantity, for

we do not exactly know how many were held—the number of

regiments required by Hadrian's scheme north of the latitude

of York amounts to about 50. This leaves only about 20 regi-

ments for the rest of the province, in which we actually know of

some 40 castella. The presumption, therefore, is that the network

of auxiliary forts which had been thrown, in varying patterns,

over northern and western Britain was now to a very consider-

able extent drained of its garrisons, which were moved up to

the Wall and its neighbourhood. And this presumption is borne

out by archaeological findings. In Wales, most of the existing

forts seem now to have been evacuated except for a military

caretaker or a skeleton garrison. The same thing certainly hap-

pened at some forts in the southern Pennine region; but our

knowledge is not detailed enough to describe the process in full.

It would be an exaggeration to say that the auxiliary troops,

which did the ordinary work of policing the country-side, were

withdrawn from every other part of Britain in order to provide

for the needs of the frontier. But the frontier absorbed two-

thirds of them; and this policy must have been based on the

assumption that the whole country behind the frontier was, if

not completely pacified, very nearly pacified, and that in future

any considerable concentration of force would be needed only

to protect the province against invasion from the unconquered

north. Danger of invasion from elsewhere, later so pressing, did

not yet exist.

THE ARMY OF BRITAIN

139

This movement of troops towards the frontier affected even the legions. The Sixth, now stationed at York, and the Twentieth at Chester, still maintained the two northern fortresses; but it was no longer thought necessary to keep the Second at Caerleon up to full strength. About the time when the Wall was built, excavation in that fortress shows traces of a decline in occupation. The head-quarters of the legion were not moved, but it seems that large drafts were more or less permanently absent. It may not be an accident that, of inscriptions connected with the building of the Wall which name a legion, by far the largest number record work done by the Second Augusta;1 nor that the only case known to us of legionaries serving in the garrison of a Wall fort is that of a draft from the same legion at Housesteads;2 and although one would expect the patrolling garrison to be composed of irregular troops, presumably consisting of the units called numeri which certainly existed on the Wall in the third century,3 there is a possibility that in the original organization of the Wall some part of this work was done with the help of men from that legion.

One curious result of this northward movement appears in Wales, where some period not earlier than the second century saw the construction of numerous massively fortified hilltop towns. These fortresses cannot have been built without the knowledge and consent of the Roman government, and it is a bold but not improbable idea that they were built at its instigation. If the government was by now sufficiently convinced of the Welsh tribes' loyalty, it might have enrolled their militia as an irregular native force for the defence of their own country, and encouraged them to fortify their dwellings. The date of this Welsh development is not very satisfactorily fixed, but it may have begun in Hadrian's reign and intensified later, when raids from Ireland became dangerous.4

1 Of inscriptions probably recording original (Hadrianic) work on the Wall, 25 belong to the Second legion, 14 to the Sixth, and 11 to the Twentieth.

2 Ephemeris Epigraphica, ix. 1177; by the style, later than Hadrian.

3 Birley, 'A note on the garrisoning of the Wall', Arch. Aeliana, ser. 4, ix (1932), p. 210.

4 Wheeler, 'Roman and Native in Wales', Y Cymmrodor, 1920–1, p. 40.

IX

THE FRONTIER AFTER HADRIAN

Begun probably in 122, Hadrian’s Wall can hardly have been finished until 127 at earliest; if in its finishing we include the replacement of turfwork by masonry as far as the Solway, a good deal later. The main part of the work was done before Platorius left Britain, which may have been about 126. After his recall we know nothing of military affairs there for some time. The province was governed, about the years 130–4, by Julius Severus, who was transferred to the east in order to suppress a Jewish revolt; the fact is good evidence for his success in Britain. It is therefore all the more puzzling to find that by 140, two years after Hadrian was succeeded by Antoninus Pius, drastic changes were being made in the British frontier-system. The general nature of these changes is clear; what is difficult to understand is their motive.

From 140 to 142 Q. Lollius Urbicus, governor of Britain, conducted a war of conquest in the Border country and the Lowlands. Reopening Agricola’s road over the Cheviots, and establishing a supply-depot at Corbridge where it crossed the Tyne, he fought his way to the Forth–Clyde isthmus, and there, at the end of the war, built a second barrier on the model of Hadrian’s Wall. Like that, the new work, known to antiquaries as the Antonine Wall, was a continuous rampart stretching from sea to sea, with a ditch in front and a series of forts, connected by a road, attached to its southern side. But it differed from Hadrian’s Wall chiefly in its simpler design and less elaborate structure.

In design it lacked two prominent features. First, there was no Vallum. Even if the explanation of the Vallum’s purpose suggested above is correct, it does not follow that there were no customs officers on the Antonine Wall, but rather that, by this time, a modus vivendi had been reached by which they could be accommodated in the forts.

Secondly, there were no milecastles and turrets, and the forts, which, apart from a few platforms apparently serving as bases for beacon-fires, were the only regular structures connected with the Wall, were on average much smaller than those of Hadrian’s system and much closer together. The Wall was only 37 miles

long instead of 73; and along its line, which is everywhere chosen

with the greatest skill so as to command a view of the valley to

northwards, there seem to have been no less than 19 forts, placed

therefore on average only 2 miles apart instead of of 4½ as on

Hadrian's barrier. In size they fall, so far as their dimensions

are known, into four groups: first comes Mumrills, in a class by

itself, at 6½ acres, with a large commandant's house suggesting

that it may have been the residence of the senior officer on the

Wall; then Balmuidy and Old Kilpatrick, both in positions of

importance and measuring 4 acres each; then a group of 5 at

about 3 acres; and lastly 4 that measure between 1 acre and 2.

This double feature, the absence of milecastles and turrets,

together with the smallness and close spacing of the forts, indi-

cates a departure from the Hadrianic model in the matter of

organization. Instead of two separate garrisons, one consisting

of whole regiments kept intact for use in the field, the other of

small units patrolling the line, there was on the Antonine Wall

only one garrison, whose units were split up so as to discharge

both functions: no doubt with consequent loss of efficiency as

fighting forces. That a system of this kind was adopted is clear

from the small size of the forts. A milliary cohort required a fort

of at least 4 acres, preferably 5; a quingenary cohort, one of at

least 2½. Castlecary on the Antonine Wall has yielded inscrip-

tions of two different milliary cohorts, one of which is known to

have had a cavalry contingent. Such a cohort can hardly have

settled down in quarters covering less than 4 acres; Castlecary

measures little over 3. Similarly, the quingenary cohort which

has recorded its presence at Rough Castle can never have lodged

in the bare acre of ground which the ramparts there enclose.

Evidently, then, the forts of the Antonine Wall were not meant

to be occupied each by a complete regiment. The units must

have been subdivided, except for a few inhabiting the largest

forts; elsewhere it must have been a frequent practice to detach

a century or two from a regiment occupying a fort of moderate

size, and thus form a garrison for one of the smallest forts. So

distributed, the total garrison of the Wall, according to careful

estimates, did not exceed six or seven thousand men.

Even this small force, however, seems not to have been wholly

composed of auxiliaries. Evidence for identifying the garrison

has been found at no more than eight forts, and in five of these

it suggests that, whether permanently or only for a time, the

142

THE FRONTIER AFTER HADRIAN

place was held by a draft from a legion or legions: in two cases the Second alone, in one the Sixth alone, in one the Second and Sixth, and in one the Second and Twentieth. If these cases are a fair sample, perhaps a third of the total garrison consisted of legionaries. This fact, taken together with the way in which auxiliary regiments were subdivided between various forts, shows convincingly enough that the garrisoning of the Antonine Wall was planned with a view to economy in men.

Economy in labour was considered no less carefully. Here again the contrast with Hadrian’s Wall is striking. The ditch that lies in front of the rampart is even larger than Hadrian’s, but the rampart itself, instead of stone, is made of turf in the western and central part, of clay in the eastern. Hadrian himself had laid it down that turfwork was very much easier to construct than masonry. And the measurements increase the contrast. The turf part of Hadrian’s Wall is twenty feet wide at the base; the Antonine Wall is only fourteen, which implies that, if the height of the turfwork was the same in the two cases,1 the Antonine rampart required, for any given length, only two-thirds of the turf that would be required by Hadrian’s. The forts, again, instead of being massively walled in stone, with monumental gateways, were surrounded for the most part with turf or earthen ramparts whose timber gateways were commonly of the simplest design; where stone was used, the construction was simple and inexpensive. Even the official central buildings in the forts were not uniformly of stone, and the barracks were of the cheapest, wooden hutments which in some cases had thatched roofs.

Both in construction and in organization, then, the Antonine Wall bears the marks of a deliberate effort after cheapness, at the cost of a serious decrease in efficiency. The same thing is to be seen if we consider its strategic position. Both its flanks, especially the left, lie unprotected upon narrow estuaries, easily crossed by the smallest craft in almost any weather. If we recollect the care with which Hadrian (in addition to maintaining ships in the Solway, as he presumably did) fortified the whole of the north-west Cumberland coast for thirty or forty miles beyond the terminus of the Wall, the complete absence of Roman posts on the Clyde below Old Kilpatrick becomes so striking that we cannot put it down to negligence. These various features of

1 See Richmond in J.R.S. xxvi (1936), p. 191.

THE FRONTIER POLICY OF PIUS

143

the Antonine Wall, when considered together, seem less like a

series of oversights than parts of a deliberate policy, based on

the assumption that a powerful frontier-work on that line was

not needed.

This impression can fortunately be verified by comparing the

Antonine Wall with the new frontier-line constructed by the

same emperor, only some five years later, in Germany. The

contrast is striking. On the ‘outer limes’ there is no subdivision

of fighting units, and no confusion between the fighting and

patrolling garrisons. Exactly as on Hadrian’s Wall, the dis-

tinction is carefully made between major forts accommodating

cohorts or alae and fortlets and signal-towers for the patrolling

garrison. And all these structures alike are solidly walled in

stone, the major forts amply large enough for their garrisons.

From every point of view the standard of strength and efficiency

displayed by the new Antonine limes in Germany is fully equal

to anything in Germany of Hadrianic date, and far superior to

that of the Antonine Wall. It is evident, then, that the new

features of the Antonine Wall are not due to any general cause,

such as a cheeseparing policy on the part of Antoninus Pius,

or a decline in the efficiency of Roman military engineering, or

a less exacting conception of what was demanded in frontier-

works. Their explanation must be sought in conditions peculiar

to Britain.

Any attempt to find such an explanation must begin from the

fact that we have to do not with a shifting of the frontier, but

with a doubling of it. Hadrian’s Wall was not evacuated. Some

of its garrisons were moved to the new line, but others were not;

and the places of those that were moved did not lie vacant. The

First Cohort of Tungrians, for example, was moved from Housesteads on the old Wall to Castlecary on the new, and an inscrip-

tion seems to show that its place was taken by a draft from the

Second legion.

In Germany, too, the new frontier created by Pius was a

double one, that is, he added a new line to run in front of one

already existing. And this was not because the tribes beyond

it were peculiarly dangerous; on the contrary, the district be-

yond it seems to have been to all intents and purposes unin-

habited. It has been suggested by the highest authority on the

German frontier that the motive of this double line was not to

keep enemies out, but to keep subjects in: to prevent certain

144

THE FRONTIER AFTER HADRIAN

barbarians, lately conquered and deported to this region, from leaving their new homes and escaping into the wilds of Germany.1

If the Antonine Wall, like the ‘outer limes’, was designed rather because of what lay behind it than because of what lay in front, its unprotected flanks are no longer a fault. And if we can find reason for thinking that the potential enemies against whom it was built were, through some special cause, unusually weak, the unusual weakness of the Wall itself becomes only reasonable. In order to pursue this line of inquiry we must go back to the time of Hadrian and consider the relation between his Wall and the natives tribes through whose territory it ran.

Hadrian’s frontier, as we saw, was established after a military disaster to consolidate what had become a dangerous position. The conception underlying it, although the Wall was not in the strict sense a fortification, was defensive: not a tactical defence, but a strategical defence. There is nowhere any suggestion that Hadrian launched a counter-attack against the northern tribes whose activities were compelling him to build his lines; on the contrary, all his available military force was thrown into building and digging. And, if he did not visit his wrath on the tribes outside the Wall, neither does he seem to have taken any severe measures against those who lived behind it. What we know of the history of native villages in Brigantia suggests that they were little, if at all, affected by Hadrian’s visit. They may have been more closely supervised by an increased force of police-troops, but there was no depopulation, no devastation. The history of native arts and crafts tells the same story.

When the north of England and the south of Scotland were conquered in the Flavian period, they had been as yet very little affected by the fashion, so richly developed in the south, of building hill-forts.2 This is sufficiently attested by the extreme rarity of such forts, to this day, in northern England south of the Wall. When we turn to the Lowlands, the contrast is startling. The highest levels, whether of the Cheviots or of the central lowland plateau, are bare of them; but all round the upper basins of the Tweed, Annan, and Clyde they come in crowds. The counties of Dumfries, Roxburgh, Selkirk, and Peebles number them by hundreds, and a dense sprinkling

1 Fabricius, in Der obergermanisch-rätische Limes, Lieferung xlviii, 53.

2 See above, ch. II, p. 23.

THE LOWLANDS AND THEIR PEOPLE

145

of them stretches over the Lammermuirs to the Forth and

the sea.

As yet little is known in detail about these forts; but the little

that is known shows that in some cases at least they were in-

habited in the second century,1 and in the circumstances this

must imply a date between the Roman evacuation of Scotland

and its reoccupation under Pius, for we can hardly imagine

them to have been built under the very eyes of the Roman

cohorts. It is likely, then, that the reigns of Trajan and Hadrian

saw a rapid growth of little forts all over the most fertile regions

of the central and eastern Lowlands; and it is even possible that

this growth was in some way conditioned by a desire to control

the lines, by now already existing, of the Roman roads. This

movement had probably begun before Hadrian’s Wall was built,

and may have been one factor in the military situation which

led to its building. If so, that building certainly did not stop it.

A development of this kind may well have caused serious

questionings. Hadrian, in the conditions in which he found

himself, was doubtless right to be content with a strategic de-

fence and refrain from anything like a counter-attack; but was

it possible to maintain that policy for ever? Were not the low-

land tribes taking advantage of their impunity to gather a force,

and establish a position, which would one day menace the safety

of the Wall itself? Was it not wiser to anticipate that day and

strike first?

Any stroke based on reasoning like this must aim at placing

in Roman hands the entire control of the lowland area. The

distribution of native forts is proof that the movement had gone

too far to be dealt with by any mere raid. Further, it is clear

that in its nature the stroke must be designed, not so much to

defeat the lowland tribes in a decisive pitched battle, as to

crush their strongholds one by one and forestall any renewal of

the movement that had created them.

At this point good fortune offers us a verification of the hypo-

thesis. One possible method of achieving this end would have

been to sprinkle the country with auxiliary garrisons as dense as

those of northern England. This Lollius did not do, and for an

excellent reason. We know already that he did not possess the

troops. Even if he completely denuded Britain of auxiliaries as far

north as the latitude of York, he would only have twenty or

1 Cf. Childe in Antiquaries Journal, xiii (1933), p. 1.

146

THE FRONTIER AFTER HADRIAN

twenty-five regiments at his disposal, and these, when the necessary

new frontier had been provided for, were not nearly enough.

A second possible method was to depopulate the country, not

completely, but sufficiently to secure the end. The brief account

which we possess of the Antonine conquest in Scotland1 men-

tions that it involved a ‘removal’ of those barbarians whom Lol-

lius conquered. The phrase, taken in itself, might be thought

ambiguous; it might conceivably2 refer, not to the transplanta-

tion of a barbarian tribe, but to the rout and flight of a barbarian

army. Yet this is not its most natural meaning; it rather sug-

gests a dislodgement of the tribes from their strongholds, in-

volving not their destruction, as Agricola had destroyed the

Ordovices, but their removal to other homes.

Here, once more, the history of the German frontier throws

light on the British.3 When the ‘outer limes’ was built and the

inner line parallel to it reorganized, about 148, a division was

made, as we have already seen, between the fighting garrison

in the major forts and the patrolling garrison in the fortlets.

This patrolling garrison, as constituted by the Antonine re-

organization, seems to have consisted entirely of numeri or small

irregular units of Britons, first apparently placed on the old

inner line, later moved forward to the outer. The number of

these units shows that there must have been, at this time or

just before it, a very large transplantation of Britons to the

Württemberg forests, not only, if Professor Fabricius has inter-

preted the evidence aright, of able-bodied men drafted into new

formations, but of entire tribes, placed in new homes where

their menfolk could be made useful and kept under super-

vision. This at once confirms and explains the statement of

Capitolinus.

The motives which led to the building of the Antonine Wall

are now clear. Hadrian’s was to remain the chief bulwark of

the province; but there was need to relieve it of a threat,

gradually becoming more and more serious, from the tribes of

the central and eastern Lowlands, which may have been in-

1 Scriptores Hist. Aug., Capitolinus, Antoninus Pius, ch. 5, § 4: ‘Britannos per Lollium Urbicum legatum vict, alio muro caespiticio submotis barbaris ducto.’

2 If taken as an echo of summotis velut in aliam insulam hostibus, Tacitus’s words

describing Agricola’s fortification of the same isthmus (Agric. xxiii), as Sir George

Macdonald (Wall, ed. 2, p. 49) suggests.

3 Fabricius, ‘Ein Limesproblem’, in Festschrift der Universität Freiburg, 1902; and

more recently in Der obergerm.-rätische Limes, Lieferung xlviii (1932), pp. 49–55.

THE PURPOSE OF THE ANTONINE WALL

147

creasing in numbers and were certainly, by building their forts

everywhere, increasing in strength. The Wall was finished, and

force was available for a counter-attack. The strategy for such

an attack had been settled once for all by Agricola; all that was

needed was to repair his roads and reoccupy the sites of his forts.

He had even driven a limes of the earlier Flavian type, a mere

chain of military posts, across the Forth-Clyde isthmus, choos-

ing their sites, as usual, so skilfully that his choice could be

accepted as final; and, in fact, almost every fort on the Antonine

Wall where any digging has been done has yielded traces of an

Agricola work beneath it. The plan of Urbicus, then, was first

to destroy the lowland native forts and deport the tribes to

which they belonged; the forts are almost all in the central and

eastern districts, so that a campaign in Galloway and Ayrshire

was not called for; and then to reoccupy the isthmus with a

limes of the new type, following Agricola once more in throwing

out advanced posts beyond it as far as Inchuthill. It was as-

sumed that the lowland tribes were the chief enemy to be dealt

with, and that central Scotland would give comparatively little

trouble. And because the Lowlands had been to a great extent

cleared of their inhabitants, it was possible to build the new

limes cheaply, hold it lightly, and support it with an unusually

thin network of forts in the rear.

On these assumptions the scheme was in itself reasonable; but

the cost must be counted. The new barrier, with its lines of

communication and outposts, however lightly held and sup-

ported, could not be manned with less than about twenty auxi-

liary regiments; it would consume almost all the balance which

Hadrian's scheme had left over for the southern Pennines and

Wales. These districts could not even now be altogether emp-

tied of troops, but their garrisons were drastically overhauled

and redistributed. One hint of such redistribution reaches us

from the Forden Gaer near Welshpool, where, by an exception

to the general rule for Wales, a more intense occupation begins

just at this time. The policy adopted was to cut down auxiliary

garrisons to the barest margin of safety, and to get the rest of

the men whom the scheme demanded by drafts from the legions,

which until then, except for possible inroads by Hadrian on the

strength of the Second, had been kept intact as mobile forces

for use in the field. In two ways the new move was a gamble:

it assumed that the central Scottish tribes would not be

148

THE FRONTIER AFTER HADRIAN

dangerous, and it assumed that, behind the frontier, Britain

was pacified.

Urbicus cannot have been unaware of these risks; and we

may wonder whether, in taking them, he meant to take them

as a temporary measure or in perpetuity. Granted the in-

security of the Antonine Wall's strategic position, the slightness

of its works, the makeshift character of its organization, and the

strain which, even so, it imposed on the resources of the province,

it may be thought to resemble a temporary measure, to last until

the pacification of the Lowlands had stood the test of time,

rather than a revision of the frontier system designed for per-

manence. In the end, as we shall see, the northern Wall was

peacefully and deliberately evacuated. If this was from the first

intended to be its end, its lack of resemblance to other frontier-

works of the same period is easier to understand. Recent excava-

tion has shown, for example, that at Mumrills, the largest of all

its forts, the commanding officer's quarters were at first of tim-

ber, a thing hardly to be expected at that time in a structure of

such importance designed to be permanent. But this is a ques-

tion which we have no means of answering.

The scheme went through without a hitch. By 142 or early

in 143 the new Wall was finished. Its lighter structures were

easier and quicker to build than those of Hadrian, and more

efficient methods of working were adopted. Instead of dividing

up the work wastefully among centuries engaged on short

sectors at a time, six powerful working-parties were made up,

two from each legion; and it seems that the entire length of the

Wall was divided into nine sectors, of which the first six, begin-

ning on the east, were simultaneously made, Wall and ditch

together, by these six gangs, which then joined forces to com-

plete the last three. The building of the forts was entrusted to

their future garrisons. Each gang signed its work by erecting,

at each end of every sector which it made, a stone slab recording

the name of the gang and the length of the sector; and it is the

text of these 'distance-slabs' that informs us how the work was

done.1

It was soon to be proved that, of the risks involved in the new

1

We owe the interpretation of their evidence to Sir George Macdonald (The

Roman Wall in Scotland, ed. 2, 1934, ch. x), who has also put forward an hypothesis

as to how this original scheme was thrown out of gear by the unexpected difficulty

of cutting the ditch through the basalt of Croy Hill.

REVOLT IN BRIGANTIA

149

frontier system, one at least was greater than had been realized. We are told by Pausanias that Antoninus Pius annexed the territory of the Brigantes because they had invaded some unidentified district which he calls the Genunian region. The statement reads as if Pausanias conceived the Brigantes as living outside the Roman frontier, whereas by now they had long been subject to Rome (it is probable that their territory did not extend far north of Hadrian's Wall, which as a military line was, of course, under no obligation punctiliously to respect tribal boundaries), and had, indeed, a capital in the Roman style at Aldborough. But sense can be made of it by supposing that what Pausanias was referring to was not a frontier war, but a revolt, and in that case it was no doubt a revolt directly due to the weakening of the garrisons in Brigantia entailed by the building of the Antonine Wall.

The revolt can be dated1 by the help of inscriptions and coins. Inscriptions show that a certain C. Julius Verus brought reinforcements from Germany for all three British legions in the time of Pius, built or rebuilt forts at Brough in Derbyshire and Birrens in Dumfriesshire, the first certainly and the second perhaps in Brigantian territory, and that his work at Birrens fell in the year 158. In that same year, another inscription tells us, repairs were carried out on the masonry of Hadrian's Wall; and excavators have noted that in various buildings of the same Wall new floors have been laid at a time which cannot be very remote from this, sometimes over layers of dark matter which might be caused by the burning of their woodwork. Coins of Pius tell us that in 155 his troops gained a considerable victory in Britain.

Many of these facts, if not all of them, must be connected with the event recorded by Pausanias. The Brigantes evidently rose about 154; the auxiliary garrisons in their territory were too few and far between to control the rising in its early stages, and the legions were too much weakened by the absence of drafts on the frontier to put it down. Verus, bringing reinforcements, suppressed the revolt in 155, and the victory was thought great enough to justify a special issue of coins. If so, the revolt may well have been serious enough to involve a certain amount of

1 As Haverfield pointed out: ‘Julius Verus, a Roman governor of Britain’, in Proc. Soc. Ant. Scot. xxxviii (1902-3), p. 454; for the coins, Macdonald, Wall (cit.), p. 10.

3720.1

L

150

THE FRONTIER AFTER HADRIAN

damage to Hadrian's Wall, taken unexpectedly in the rear, and

the destruction of some forts immediately to the north of it.

The trouble may even have spread as far as the Antonine

Wall. Buildings in the forts there show signs of having been

twice destroyed before being finally abandoned. Neither de-

struction can be accurately dated by archaeological or indeed

any other evidence; but it is a reasonable conjecture that the

first of them may have happened at or about the time of the

Brigantian revolt. If this conjecture is correct, the inference

follows that not one only, but both of the risks entailed by

Lollius's distribution of his forces had been under-estimated.

Nevertheless, no radical change was made. Apart from details

here and there, too scanty to serve as the basis for a theory, we

do not know on what principles the rebuilding of the Antonine

Wall forts was planned. At Mumrills the wooden house for the

commanding officer seems to have been handsomely rebuilt in

stone; at Cadder, on the contrary, a stone house was replaced

by a timber one. At Rough Castle the rampart was strengthened

by adding a new rampart-walk, and it is thought that the size

of the garrison was reduced. No generalization can be based

on details like these. Nor can we even be sure how far Julius

Verus strengthened the auxiliary garrisons of Brigantia, and

how far he merely rebuilt their forts and restored the status quo.

In either case, he was only patching a system whose weaknesses

were fundamental.

It is therefore not surprising to learn that within a very few

years trouble again broke out in Britain. Verus was succeeded

as governor by M. Statius Priscus, who was there when Pius

died in 161 and was succeeded by Marcus Aurelius. The new

reign opened darkly. The Historia Augusta enumerates a severe

flooding of the Tiber, a war in Parthia and the rout of the

Roman governor of Syria, an invasion of unconquered Germans

into the provinces of Germany and Raetia, and a British war.

This last was dealt with by sending out a new governor, Sextus

Calpurnius Agricola, doubtless to succeed Priscus, under whom

the trouble must have arisen. Of its nature we know nothing.

There are traces of his rule on Hadrian's Wall: a building dedi-

cated to the 'Invincible Sun-God' at Corbridge, an altar to the

'Syrian Goddess' near Carvoran, a building at Ribchester, were

set up while he was governor, or even at his express command;

but whatever temporary success he may have enjoyed, war broke

DISASTER UNDER COMMODUS

151

out again, unless the Historia Augusta is hopelessly confused, in

169. Another and more trustworthy writer tells us that in 175

Marcus sent to Britain 5,500 auxiliary cavalry. He was evidently

aware that so long as the present frontier system there was in

being the army of Britain urgently needed reinforcement; but

his wars on the Danube forbade more effective action.1

Matters came to a head in the reign of Commodus, who suc-

ceeded his father as sole emperor in March 180. The last war

of which we know anything definite began within the frontier;

this time it came from outside. The tribes of central Scotland,

encouraged no doubt by the feeble defensive power of the

Antonine Wall, attacked it in force and swept across it into the

Lowlands. It was a critical moment in the history of the British

frontier, not only as the gravest blow it had ever yet suffered

since its first definite organization by Hadrian, but as the death-

knell of an age. Hitherto the initiative had lain with Rome.

Henceforth it lay with her enemies. The barbarians were never

again to forget the triumphs and the plunder that awaited them

beyond those barriers. Rome, however successfully she main-

tained her defences in the future, was never again to win a

decisive victory. With this invasion began the long process

which, checked again and again, was never reversed until the

helpless Britons of two centuries and a half later wrote their

lamentation to Aetius: ‘The barbarians drive us into the sea,

the sea drives us back to the barbarians.’

The exact date of the war is not known. Cassius Dio gives

an account of it under the year 184; that, however, was the

year of its ending, the year when Commodus took the title

Britannicus and issued coins commemorating a British victory.

Dio’s account has reached us only in an abridgement, but it is,

as usual, far superior to the vague generalities of the Historia

Augusta:[Of the wars waged by Commodus] the greatest was

in Britain. The tribes in that island crossed the wall which

1 Scriptores Hist. Aug., Capitolinus, M. Antoninus, ch. 8, § 7, ch. 22, § 1. Cassius

Dio, lxxi. 16. If we could be sure that the archaeologists are right in now dating

the first destruction of the Antonine Wall to c. 154 and the second to 180–3, an

interesting result would follow in connexion with these British disturbances under

Marcus: since on that view the Antonine Wall was unaffected by them, they must

have been internal revolts, and thus testify to incomplete pacification of the pro-

vince. I have not drawn this inference in the text because the details of the

chronological scheme for the Antonine Wall periods are not, in my opinion,

sufficiently certain to warrant it: there is really nothing to prove that the first

destruction did not happen in 161–2.

152

THE FRONTIER AFTER HADRIAN

divides them from the Roman fortresses, and did great harm.

They slew a (Roman) general and the men under his command.

Commodus, greatly alarmed, sent Ulpius Marcellus against

them.'1 A description of Ulpius follows: a hard man, sparing

and austere in his own life, relentless in his demands upon

others; fanatical in his refusal of food and sleep, and strangely

ingenious to invent new disciplines for his army. ‘Such then

was this Marcellus; who inflicted a terrible defeat on the bar-

barians of Britain.’

From this passage it is evident that the war began not as a

revolt, but as an invasion. The tribes beyond the Wall broke in

and destroyed a considerable force, apparently commanded by

the governor in person. A new governor was sent out, who

repulsed the invaders. Ulpius Marcellus, as an inscription from

Hadrian’s Wall tells us, was already governor of Britain in the

joint reign of Marcus and Commodus (177-80); Dio’s words

imply that the war broke out under his successor, and that

Ulpius was sent back to retrieve the disaster.

Only one Wall is mentioned, and that in terms which might

apply to either; for ‘the Roman fortresses’ here are the per-

manent quarters of the legions. But Hadrian’s Wall, according

to the findings of archaeology, suffered no disaster at this time;2

and it is easiest to identify Dio’s Wall with the Antonine barrier,

and to see traces of the invasion in the second of the two destruc-

tions which, as we already know, overtook the buildings of that

Wall in the course of their history. On this assumption, the

story of the war is that the tribes of central Scotland over-

whelmed the Antonine Wall, ravaged the Lowlands, met and

destroyed a considerable force sent to stop them, but failed to

break the much more solid barrier of Hadrian. The field army,

beaten and demoralized, was brought up to fighting-pitch again

by the stern vigour of Marcellus, and attacked the invaders once

more, this time successfully. An inscription at Carlisle records

the dedication of a shrine ‘to the divine companions of the god

1 Cassius Dio (Xiphiline’s abridgement), lxxii. 8. Hist. Aug., Lampridius,

Commodus, ch. 13, § 5, speaks vaguely of a revolt in Britain, and ibid., ch. 8, § 4,

seems to confuse this war with the soldiers’ rebellion to be mentioned below.

2 Until lately it was assumed that a great destruction everywhere traceable on

that Wall could be brought into connexion with this passage of Dio. The conse-

quences of this assumption were, however, difficult to accept; and within recent

years new discoveries have shown that the destruction actually took place later,

on the occasion of Albinus’s removing his troops to Gaul.

THE END OF THE ANTONINE WALL

153

Hercules', with whom Commodus was pleased to identify him-

self, 'on the rout of a vast horde of barbarians'; Professor

Rostovtzeff, who has restored the inscription, sees in it a com-

memoration of Marcellus's victory.1

If that is the story of the campaign, its lessons were not diffi-

cult to read. The Antonine Wall had proved powerless against

an attack from the north, and it had now become evident that

the possibility of such attacks was a thing to be reckoned with.

The position in Scotland was untenable. Hadrian's Wall, on

the contrary, had signally justified itself. It had not only

checked the barbarians, it had given the shattered legions time

to recover their morale behind its shelter and prepare their

counter-stroke at leisure.

If the latest reading of the evidence is right,2 this lesson was

understood and its practical consequences drawn by Marcellus

himself. The third and last period of construction on the

Antonine Wall is now regarded as his work; but the buildings

thus reconstructed seem not to have been long occupied; and

there is much to show that when they were finally abandoned,

which in any case happened before the end of the century, they

were methodically dismantled not by triumphant enemies, but

by the Romans themselves, for altars and inscribed slabs, which

enemies would take special pleasure in destroying, were hidden

out of harm's way in the ground. All this suggests that when

Marcellus had defeated the invaders his first act was to repair

the Antonine Wall, as a demonstration of strength, and that

then, when order was restored and the frontier quiet, he

1 C.I.L. vii. 924; J.R.S. xiii (1923), p. 91.

2 Sir George Macdonald's, in Wall (cit.), ed. 2. After the first scientifically

conducted excavations on the Antonine Wall, the evidence of the coins found there

was interpreted as indicating that the Wall was finally abandoned when the in-

vaders of 180–4 destroyed it: the first intermediate destruction being assumed to

date c. 155, the second was associated 'with the operations of Calpurnius Agricola

or with a later rising in the reign of Marcus' (op. cit., ed. 1, 1911, p. 401). Sir

George now thinks the interval of time thus allowed for, between the first and

second destructions, insufficient to account for the state in which remains of the

second period have been found (ed. 2, p. 479), and therefore dates the second

destruction to 180–4 and ascribes the third period to Ulpius Marcellus himself.

This reopens the question, when was the Antonine Wall finally evacuated? Sir

George, on an exhaustive review of the evidence, comes to the conclusion that this

event took place early in the reign of Commodus and was in fact carried out by

Marcellus in an orderly and deliberate way. The evidence does not amount, in

my opinion, to proof, though this is not the place to go into details; none the less,

I am disposed to think that on the balance of probabilities this view is the right

one, and adopt it in the text.

154

THE FRONTIER AFTER HADRIAN

deliberately razed the fortifications and evacuated the position,

withdrew to Hadrian’s line, and made that for the future the

single frontier of Roman Britain.

It must have been soon after this event that Marcellus was

recalled. We are told that Commodus had it in mind to make

away with him on account of his virtues, but finally spared him.

Perhaps the emperor was angry with the man who had suffi-

cient strength of mind to abandon the Antonine Wall; perhaps,

on the other hand, he accused Marcellus of driving his troops to

mutiny by the unreasonable severity of his discipline. For it is

certain that, when Marcellus was succeeded by Helvius Perti-

nax, the future emperor, the army of Britain was in a dangerous

mood, airing grievances and ready for violence. The story of

these mutinous moods is a confused one, as such stories generally

are. At one time the men attempted to set up an emperor of

their own, one Priscus; at another, they took it into their heads

to send a deputation, 1,500 strong, to wait upon Commodus and

denounce his trusted minister, Tigidius Perennis, as a traitor.

The idea was not an invention of their own. Perennis had a

son who was in command of the Illyrian army; and certain

Illyrian soldiers in Rome went to the emperor and informed him

that Perennis meant to put this son on the throne in his stead.

Apparently the deputation from Britain went to back up the

statement of these Illyrians. In any case they gained their

point; Perennis lost at once his power and his head. His fall

took place in 185, and it was after this event that Pertinax was

sent to restore order in Britain. He put down the mutiny, not

without danger to his own life, when once more the mood of

the troops abruptly changed, and they declared themselves

willing to serve whatever emperor he liked, preferably himself.

The real importance of these outbreaks lies in the fact that

they foretold a new civil war. The assassination of Commodus

in 193 brought about a situation much like that following the

death of Nero in 69. In both cases the army took upon itself

the nomination of a successor. In both cases the praetorian

guards were the first in the field; in both cases their choice was

a failure, and the matter was left to the arbitrament of the

sword, each army fighting the rest in support of its own man.

The praetorians chose Pertinax, but he alienated his supporters

by incautious reforms, and the soldiers who had crowned him

butchered him, and proceeded to hold an auction, offering the

empire to the highest bidder. So contemptible a farce could

only end contemptibly. Didius Julianus, to whom the majesty

and godhead of an emperor were knocked down for cash, paid

his money but never received his goods: he remained a futile

and incompetent old man, waiting in fussy inactivity while

Septimius Severus, the grim African who commanded the

legions of upper Pannonia, made all ready for his destruction.

By July Julianus was dead and Severus was in Rome. His

first act was to disband the praetorian guards and substitute a

bodyguard of his own choosing. He knew that Pescennius Niger

in Syria and Clodius Albinus, the successor of Pertinax in

Britain, were possible rivals, and he preferred to deal with them

singly. He conciliated Albinus by giving him the title of Caesar,

virtually appointing him as his own successor, and marched to

the east, where he crushed Niger and fought various frontier

wars. The story of his breach with Albinus is not altogether

clear; but evidently neither believed that a breach could be

avoided, and each was manœuvring for position when it should

come. Correspondence treasonable to Severus passed between

Albinus and the senate: it came to the emperor's hands, and

he declared Albinus a public enemy. In reply, Albinus had

himself proclaimed emperor, crossed with all the troops he

could muster into Gaul, and set up his court at Lyons. Septimius

was already in Pannonia, and by February 196 his army con-

fronted that of Albinus outside Lyons. Albinus, decisively

beaten, committed suicide.

Albinus's bid for the throne was a disaster to Britain, not

because it was made in vain, but because it was made at all.

There can be no doubt that, the Antonine Wall once given up,

Hadrian's frontier was equal to its work. Nothing that the

tribes beyond it could do at this time need have caused its de-

fenders a moment's uneasiness. In fact, so far as our knowledge

goes, down to the very end of its history Hadrian's Wall never

fell before a frontal attack. It was captured by the enemies of

Rome only when its garrison was either withdrawn or else in

league with those same enemies. But now, not many years after

one successful invasion, the door was deliberately thrown open

to another by Albinus's removal of the garrisons to fight for him

in Gaul.

The consequence has long been known to archaeologists. All

along Hadrian's Wall forts, milecastles, turrets, and the very

156

THE FRONTIER AFTER HADRIAN

fabric of the rampart itself, have been towards the end of the

second century wrecked with astonishing thoroughness. It was

no mere question of burning whatever would burn. At a fort

like Housesteads visitors can still see how the great squared

stones of the gateways have been forcibly levered sideways out

of place, down to the lowest course. At the neighbouring mile-

castle the piers of the north gate have been forced bodily away

from the walls with such violence that large stones bonding them

into the walls had been snapped in two, and the flags of the

floor, when the milecastle was first dug, were found to have

been tilted up on edge with the impact of masonry falling on

them. Elsewhere, the same tale is repeated over and over again,

how the debris of this destruction is full of stones from the ruined

masonry of the buildings in which it lies. In many places the

Wall itself has been dismantled to the very foundations. Nor is

all this peculiar to the Wall. It happens all over the north.

Even at the legionary fortress of York the foundations of the

stone rampart put up in Trajan's time have been deliberately

undercut and the wall thrown down, length by length. At

Chester Severus's men had to rebuild the walls of the fortress

from ground-level.

This flood of destruction which swept over northern Britain

as far south as the latitude of York and Chester cannot be accur-

ately dated by archaeology alone. All that can tell us is that

its effects were repaired by Severus, beginning at least as early

as 198; and we may be sure that no government with the

smallest degree of self-respect (a quality lacking neither in

Ulpins Marcellus, Pertinax, nor Albinus) would have allowed

its fortifications to lie thus ruined for a day longer than was

necessary. Archaeology further tells us three things about the

damage itself. First, it was plainly the work of men at leisure.

It was not done in the heat of a siege; it was done in a cold fury

of destruction by men unopposed and uninterrupted at their

work. Secondly, it did not spread to the walled towns of civil-

ized Britain. Thirdly, unlike the destruction of the Antonine

Wall, it was not done by the garrisons themselves, intending a

final abandonment of their fortifications, for it is not associated

with deliberate concealment of altars and other things that in

such a case would have been piously preserved from outrage.1

1 There is one interesting exception to this rule: the great hoard, so to call it,

of altars buried at Maryport and discovered accidentally in 1870. But against this

INVASION BY THE MAEATAE

157

One of the first acts of Severus, when his victory was complete, was to send to Britain a new governor, Virius Lupus. It is recorded that when Lupus reached Britain he found it overrun by the Maeatae, a tribe or confederacy of Scotland; and so firm was their hold on some considerable part at least of the province that, unable to drive them out by force, and learning that the Caledonians, who lived farther to the north, were likely before long to join them, he bribed them to be gone.1 The inference can hardly be doubtful. These Maeatae were descended from the remnant of some of the tribes whom Urbicus had deported. A dozen years before they had watched Marcellus's men destroying the forts of the Antonine Wall; and when Albinus withdrew the garrisons from northern England they realized that the time had come to avenge the wrongs of their forefathers. They swept in, no one resisting them, and deliberately treated every Roman building they could find in the same manner, arguing that without their fortifications the Romans could do nothing. If they did not destroy the walls of towns, it was because their inhabitants could beat them off. They were not prepared for siege-work, and could destroy only where no opposition was offered.

Once the country was clear of them, Lupus made it his first business to rebuild the forts they had wrecked. It was he, no doubt, who robbed the cemetery at Chester for stone to reconstruct the north wall of the legionary fortress, whose demolition in the eighties gave us the finest collection of Roman tombstones in the country; he, too, that reconstructed the walls of York. But at these sites there is no inscription to identify him as the rebuilder. For that we must go first to Ilkley, where he restored a building in 197 or early in 198; Bowes, also in Yorkshire, where he rebuilt the regimental bath-house before the middle of 198; and perhaps Brough-under-Stainmore, the next fort west of Bowes, where a scarcely legible stone records work done

must be set the numerous cases of the opposite kind: second-century monuments thrown down and smashed and now existing only in fragments. And it is curious how few building inscriptions of Hadrian's own time from Hadrian's Wall have survived even in this fragmentary state.

1 The Maeatae are described as living 'near the Wall that bisects the island' (Dio, lxxvi. 12). In view of the fact that elsewhere Dio, mentioning one Wall only, means the Antonine Wall (above, p. 151), the Antonine Wall is probably meant, which places the tribe on or near the Forth-Clyde isthmus. For Lupus and the Macatae, Dio, lxxv. 5. The Caledonians, according to Ptolemy, lived in the central highlands of Perthshire and Inverness-shire, south-east of Glen More.

158

THE FRONTIER AFTER HADRIAN

apparently about the same time. Evidently Lupus was hard at

work in Brigantia. He must also, surely, have begun the enor-

mous task of repairing Hadrian’s Wall.

Later historians have handed down to us a persistent tradi-

tion that made Severus the actual builder of the Wall. Until

it was known how thoroughly that masterpiece of Hadrian’s

had been ruined by the Maeatae, the existence of such a tradi-

tion was hard to understand; and in fact it misled generations

of scholars, and for long divided opinion into two camps accord-

ing as Hadrian or Severus was supposed to be the builder. But

if the Wall itself was razed to ground-level by the mile together,

and its attendant works shattered often to their foundations, the

work of Severus was so extensive that the tradition is not ill

founded and the credit which it gives him not undeserved.

Unfortunately we know hardly anything of how this great work

of restoration was carried out, or at what speed. By 205–8 it

was far enough advanced to have reached the internal buildings

of the forts, and even to have extended north of the line to an

outpost such as Risingham; and inscriptions tell us that the

tribal communities of romanized Britain built certain sectors

of the great rampart, no doubt at their own expense.1 All this

indicates a vast effort of reconstruction, extending from 197 to

208, in which every nerve was strained and every source of

labour tapped in order to make the frontier system, as it had

been at the end of Hadrian’s reign, a going concern once more.

To do this, it was not necessary for Severus to visit Britain in

person. The system had been designed once for all by Hadrian;

to reconstitute it was work that any intelligent subordinate

could do; and it was done successively by Virius Lupus, Alfenus

Senecio, and one or two other governors coming between them.

But Hadrian’s work had been incomplete without the offensive

against Scotland that followed it under Antoninus Pius. Severus,

therefore, in forming his plans for the treatment of the British

frontier, was not content merely to restore the Hadrianic forti-

fications; he meant also to emulate Antoninus Pius in over-

running and depopulating as much as should prove necessary of

Scotland, though without repeating his mistake of creating a

second, and indefensible, frontier-line. It is clear that his policy

was, structurally and topographically, a return to Hadrian, but

1 Not necessarily of their own free will; perhaps rather by imperial orders. For

burdens imposed by Severus on towns, cf. p. 203.

FRONTIER-POLICY OF SEVERUS

159

politically, a partial return to Pius. First the fortifications were

to be built, then, when the troops had finished that task and

could be spared for a different one, the country north of them

was to be ravaged.

In 208 the first part of the programme was finished, and

Severus came over to conduct the second part himself. From

208 to 211 the emperor, now ageing and losing strength year

by year, pressed home a series of campaigns into the heart of

Scotland. Like Agricola and Lollius before him, he used the

road that crosses the Cheviot at Makendon and the Tweed at

Newstead; he created afresh a base at Corbridge, and a naval

station at Cramond on the Forth; and by immense labours of

road-building and forest-clearing he is said to have advanced

almost to the northern extremity of Britain. And, in fact, there

are Roman camps, which may well be his, as far north as Aber-

deen and beyond.

The impression which the narrative of these campaigns, as

given by Dio and Herodian, leaves on a reader's mind is one

of great efforts and heavy casualties fruitlessly borne, in blindly

striking at an enemy who could always strike back but never

stood still to be hit: guerrilla warfare skilfully waged by light-

armed warriors and charioteers, who had learnt the lesson of

avoiding a pitched battle and confined themselves to hampering

the movements of Severus in every possible way and wearing

him out by ceaseless attrition of men and material. And these

at any rate are facts: that Severus died, worn out by his labours,

at York in 211, and that his sons forthwith broke off the Cale-

donian war and returned to Rome.

It looks like the story of a misguided and wholly unsuccessful

war. Yet after its conclusion the British frontier enjoyed un-

broken peace for nearly a hundred years. If the war ended in

complete failure for Rome, why did the Scottish tribes do

nothing to follow up their victory? In part, that was due, no

doubt, to the excellence of the Hadrianic frontier and to the

efficient way in which Severus's lieutenants had restored it. But

it is difficult to say that the second phase of his frontier policy

was a failure unless we are first sure what it was meant to

achieve. Certainly not the complete conquest of Caledonia; if

that had been its aim the elaborate reconstruction of Hadrian's

Wall would have been sheer waste. Certainly not the restoration

of the Antonine Wall, or he would have got to work on that,

160

THE FRONTIER AFTER HADRIAN

as he did on restoring Hadrian's, as soon as he had possession of the ground on which it stood. In fact he never even temporarily reoccupied the sites of its forts, although many of them lay on his lines of communication.1 It seems that he did not intend a permanent occupation of any part of Scotland; the forts which, according to Dio, Caracalla evacuated at the end

of the war need not be more than a few places like Cramond, meant to be held only for its duration. The evidence, such as it is, suggests that his campaigns were meant as wars of devastation, designed to inflict the heaviest possible losses on the tribes that had risen during the absence of Albinus: punitive expeditions, visiting the wrath of Rome on enemies outside her grasp but not outside her reach. In the course of them he seems to have followed the example of Lollius Urbicus in deporting captives to the German frontier: at Walldürn there was a corps of Brittones gentiles which first appears in the third century and is thought to owe its origin to the campaigns of Severus.2 If this was his aim, to prove that Rome could strike deeply beyond her own frontier, punish those who had attacked it, and carry them off into captivity, his success may have fallen short of his desire, but it was not negligible; and the lesson may have contributed much to the security which Britain enjoyed in the generations that followed.

1 Sir G. Macdonald, Wall (cit.), p. 480, note, observes 'there is no scarcity of coins of Severus and his family in Scotland, but none has ever been found on the Forth and Clyde line, or indeed in any castellum save Cramond', where no less than nine have turned up.

2 O.R.L. Lieferung xxi, p. 13; Lieferung xlviii, p. 52.

Postscript. Fellow students (e.g. Richmond, J.R.S. xxvi (1936), pp. 190-4; and some others by letter) have not much liked what I say about the Antonine Wall. It may be of use to repeat here the logical framework underlying it, so as to clarify the issue. (1) The A. W. fell disastrously, once at least in fair fight, twice in forty years (this is common ground). (2) It was therefore much too weak for the work it was called upon to do (this is the same thing in other words). (3) But this weakness was not typical of Roman limites (cf. Hadrian's). (4) It is therefore exceptional, and must be treated as such. (5) The questions thus arise: (a) Wherein did this weakness consist? (b) Why was it tolerated? (c) Why, in spite of it, was the A. W. created at all?—These are the questions I have tried to answer; I hope it will now be clear why I felt obliged to ask them.

BOOK III

BRITAIN UNDER ROMAN RULE

X

THE MACHINERY OF GOVERNMENT

ONE feature of the settlement by which Augustus established himself in the new office of princeps was a division of the provinces into two classes, one to be controlled by himself, the other by the senate. Senatorial provinces were to be governed as they had been during the Republican age, by men of consular or praetorian rank, chosen by lot from the senior members of their class; imperial provinces were in theory all alike governed by the emperor himself as holder of the proconsulare imperium, but actually by legates of his appointing, each of whom ruled in his own province unless the emperor happened to be there. The full title of such a legate was legatus Augusti pro praetore, that is, in his capacity of governor he ranked as a practor, in order that the emperor’s proconsular power should be superior to his own; but in provinces garrisoned by more than one legion the governor was always a man of consular rank. This was the case in Britain, which, as possessing either four or three legions, was reckoned one of the most important of the imperial provinces. The emperor’s representative in charge of a province like this, with his miltary command, his ius gladii, and his five fasces, was a man invested with less dignity than the governor of a senatorial province ruled by a proconsul, but with more power; it was important that he should be carefully chosen both for ability and for loyalty, and in order that he should learn to know his province well but should not become too powerful it was the usual practice of emperors to retain him in office for a term of not less than three years and not more than five.

Among the subordinates of the propraetor the chief were, on the military side, the officers commanding legions. These, like himself, were legati Augusti, delegates and nominees of the emperor, responsible therefore to the emperor, but subordinate to the governor as commander-in-chief of the military forces of the

162

THE MACHINERY OF GOVERNMENT

province. Normally a legatus legionis accepted without question

the authority of the propraetor; but occasions were not unknown

on which a weak governor failed to command his obedience, and

we have already seen, in the case of Trebellius Maximus, how

such failure might arise, and that it might put the governor in

an untenable position.

On the judicial side a governor might have a legatus iuridicus

as his assistant. Such an office does not seem to have been com-

mon; elsewhere we know of it only in Spain and Upper

Pannonia; but in Britain the names of no less than five legati

iuridici are known, of whom the first occupied that position in

the governorship of Agricola, and the last in the early years

of the third century. It is not known whether the office came

into being under the Flavians, as some historians conjecture;

but it seems that such an official was a regular part of the

British governor's staff from at least the Flavian period to

at least that of Severus.1 Like the legates of legions, the juridical

legate was directly nominated by the emperor, but it does not

follow that he relieved the propraetor of responsibility for juris-

diction; he seems rather to have been a subordinate, appointed

to assist the hard-worked governor of a large and exacting pro-

vince in the performance of one part of his duties.

A subordinate of either kind was capable of governing a

province in the absence of the propraetor. We have seen how,

during the civil wars following Nero's death, the three legionary

legates governed Britain in the interval between Trebellius

Maximus and Vettius Bolanus. Similarly, we learn from an

inscription that about the beginning of the third century M.

Antius Crescens Calpurnianus, juridical legate, governed in the

absence of the then propraetor.

The same could be done, it would seem, though here the

evidence is less decisive, by an official of a very different type,

namely, the procurator. In theory the procurator was simply

the emperor's steward; in practice, he was the head of the

financial service in an imperial province. He was not a man of

senatorial rank, and his relation to the governor was somewhat

ambiguous. He was not under the governor's orders, but the

1 Gsell, Essai sur le principat de Domitien, pp. 140-1, conjectures that C. Salvius

Liberalis Nonius Bassus, who was legatus iuridicus in Britain c. A.D. 79, was the first

and that the office was, for Britain, created by Vespasian. Domaszewski, Rhein.

Mus. xlvi (1891), pp. 599 sqq., would place his appointment later, under Domitian.

THE FINANCIAL SERVICE

163

head of an independent department and responsible directly

to the emperor; consequently, in spite of his inferior dignity and

powers, he was capable of acting independently, and the actual

relations between him and the governor depended partly on

their respective characters and partly on the orders each re-

ceived from their master. The emperor might, and often did,

use his procurator as a check, or even as a spy, on the activities

of his governor, for the procurator’s humbler position and entire

dependence on the emperor’s favour would tend to make him

a faithful servant; and conversely, a governor might sometimes

use his influence, though he could not use the method of direct

command, to check a procurator whose rapacity was in his

opinion oppressive to the inhabitants of his province. The

events preceding and following Boudicca’s rebellion are typical

of what might happen when relations were strained; and in the

crisis of the rebellion itself the colonists of Colchester, it will be

remembered, appealed to the procurator, the governor being

out of reach, for armed protection, though the personnel at his

command was inadequate to provide what was needed.

The business of this financial department may be grouped

under various heads. The chief source of revenue was the land-

tax, based on a survey dividing land into private property,

communal property, and state property, and classifying it ac-

cording to its productivity. In addition to this and the poll-tax

paid by a conquered population, there were levies in kind, the

annona, paid in grain and devoted to the maintenance of the

military establishment. A good governor might do much for

the welfare of his provincials by checking abuses here; Agricola,

for example, we are told, found that the Britons were being

compelled to buy grain from the Roman stores to meet their

obligations, and to carry their produce long distances for delivery

when there were fortresses near at hand. There is in existence

a modius or corn-measure of Domitian’s reign, found close to

Hadrian’s Wall, certified as measuring 17½ sextarii but actually

holding about 20; it has been thought that this may have been

dishonestly used to obtain 15 per cent. more annona corn than

was justly due.1

The procurator was also in charge of imperial property,

whether public or personal to the emperor; such property in-

1 Tacitus, Agric. xix; Haverfield on the ‘Modius Claytonensis’, in Arch. Aeliana,

ser. 3, xiii. 85.

164 THE MACHINERY OF GOVERNMENT

cluded domain lands and also mines, quarries, and certain industrial establishments. Indirect taxes too came under his department; of these the most important were the customs duties, portoria, levied on all goods passing the imperial frontier. There were many other taxes, not all of which need be here mentioned: the five per cent. legacy duty was not at first applied in the provinces, being levied only on Roman citizens, until Caracalla extended the citizenship to all provincials, but they were liable to imposts on trade, on purchase of slaves, and so forth, and in especial to the aurum coronarium, an extraordinary levy, theoretically payable on the accession of a new emperor, but sometimes demanded on other occasions.

As to the details of this financial system in its application to Britain, we know little or nothing. Tacitus records, doubtless on Agricola's word, that the Britons were good tax-payers unless exasperated by official insolence. Inscriptions add over a dozen to the two names of procurators that Tacitus has preserved. A writing-tablet has lately been found in London, branded on the back with the legend 'issued by the procurators of the province of Britain', and a number of stamped tiles, also found in London, testify to the existence there of an official building belonging certainly to some body of financial officers, though their exact identification is in dispute. Taken in conjunction with the tablet, however, they can be most reasonably attributed to the procurators, and used as evidence that London (the chief centre of trade, and as we have seen originally the military arsenal) was the seat of financial administration. The totals of income and expenditure for the province are impossible to conjecture; but with its enormous military establishment we can believe Appian when he says that as a whole it did not pay its way.

Within the general framework of a Roman province there was plenty of room for local self-government. In the first place, there were institutions of strictly Roman character. Of these the most important was the colonia, which was a settlement of time-expired legionaries on the land. Its members were there- fore Roman citizens. They lived in a town and held allotments outside it, so that culturally they enjoyed an urban life and all its institutions, while economically they had the status of inde- pendent farmers. It is true that the title of colony could be conferred in an honorary sense on communities very different

from this in origin, but so far as we know it never was so con-

ferred in Britain; Tacitus goes out of his way to remark that it

had not been given to London, in spite of its size, wealth, and

importance as a commercial centre and military depot. A

colony received a charter conferring upon it, among other rights,

that of electing four annual magistrates, duoviri iuri dicundo for

jurisdiction and duoviri aediles for building and finance; and it

was to some extent free from interference by the provincial

governor in its affairs. In Britain we know of four colonies:

Colchester, founded A.D. 50, Lincoln, founded under the

Flavians, Gloucester, under Nerva (96–7), and York, perhaps

under Antoninus Pius (see p. 171).

The municipium, like the colony, was a community of Roman

citizens, or if not of full citizens at least of persons enjoying the

so-called Latin franchise. The essential difference between it

and a colony was that whereas a colony was an offshoot of the

Roman state from within, a municipium was a community taken

into the state from without. Like a colony, it received a definite

constitution; its magistrates, called quattuorviri, corresponded to

the two pairs of duoviri in a colony; but in the case of a munici-

pium this constitution was conferred by grant upon a com-

munity already existing. In Britain we have only one known

example, that of Verulam (cf. p. 86).

Towns of either type had a constitution modelled on that of

Rome herself. There was a quasi-senate, called ordo, whose

members were called decurions, and were ex-magistrates. It

was a permanent executive body which dealt with the ordinary

business of the town by decree, and was consulted on all matters

of importance by the magistrates. The magistrates were elected

by the general citizen body; but apart from this the citizens as

such had no voice in public affairs, which in practice were con-

trolled by the ordo. Every such town had a college of Augustales,

existing for the purpose of maintaining the emperor-worship

which was everywhere carefully fostered as an institution pro-

moting the spiritual unity of the empire; the Seviri Augustales

appear to have been the annual magistrates of this college, and

thus the chief religious officials of the community. A greater or

less extent of land round a colony or municipium was ‘attri-

buted’ to it; the inhabitants paid taxes to the town and could

plead in its courts and serve in its militia, though they were

not citizens.

3720.1

M

166

THE MACHINERY OF GOVERNMENT

Self-government of a simpler and less highly organized kind extended far beyond the limits of these peculiarly privileged cities. Even the smallest communities which could in any sense be called towns reproduced to some extent the same constitutional pattern. Where a group of houses grew up outside a military fort, as soon as it attained sufficient importance (and the evidence shows that no very high standard was imposed) it was recognized as a res publica classified under the title of vicus, which means either a subdivision of a town or a separate village; given a body of councillors, called vicani; and allowed to elect its own magistrates, two magistri or vicomagistri, and two aediles. In this way, whenever we find in a province such as Britain anything in the least resembling a town, even so humble an affair as the cluster of houses beside a fort often is, we can be sure that it enjoyed, possibly within very narrow limits, the right of managing its own affairs, settling its own disputes, raising and spending money, and so forth.

All this applies to urban communities either Roman in origin or (in theory at least) completely romanized. But the same flexibility which caused self-governing institutions to spring up on a scale adapted to their environment, wherever anything existed to which the name of Roman or romanized urban community could be applied, caused the same principles to be extended, so far as that was found practicable, to the native population itself. The method of bringing about this extension was to take over the political system of the natives as it stood, and modify it in the direction of conformity with the same universal model. Britain, before the Romans came, was a country whose political life was the life of a number of independent tribes, mostly ruled by kings, and so related among themselves that smaller tribal units could be absorbed into larger ones without altogether losing their individuality. There is no quite certain trace in Britain of the pagi into which the tribes of Gaul were subdivided;1 and the process of replacing monarchical by republican government, which had already gone a long way in the Gaul of Caesar’s time, does not seem to have gone far in Britain by the time of the Roman conquest. In Gaul, outside the ancient Province of the south, the Romans retained this tribal system as a weapon of local government.

1 There is a possible trace: the curia Textoverdorurn of C.I.L. vii. 712: cf. Stevens, Arch. Aeliana, ser. 4, xi (1934), p. 138.

TRIBAL SELF-GOVERNMENT

167

The tribal aristocracies easily became romanized in manners and thought; they were induced to build towns, a capital city for each tribe, in the Roman style; and in each of these towns the native republican institutions were readily modified, with little more than verbal alteration, into an executive ordo and annual magistrates. By the time Britain was conquered a precedent was thus well established for retaining the tribal units and making them into units for a local self-government rooted in native tradition and at the same time conformable to Roman usage. We have already seen how far Roman conservatism went in this respect, by using even the kings of these tribes, when they were willing to be so used, as imperial legates. In a town like Noviomagus of the Regni (Chichester) there must have been, for a generation after the conquest, a complete microcosm of the Roman constitution: the king Cogidubnus in the position of vice-emperor, an ordo of local notables representing the senate, and annually elected officers corresponding to the magistrates of the imperial city. But we have also seen that this was a temporary expedient, adopted only to ease the transition from barbarian freedom to the full membership of the Roman commonwealth which Tacitus called servitude. Ultimately, all these tribal units settled down into the standard constitutional pattern of a municipality, differing mainly in the legal status of their members; for these civitates (in a British context the word practically means tribes) did not enjoy Roman or even Latin citizen-rights.

This system was applied to the greater part of Britain, at least in the lowland zone. But its application involved a certain degree of artificiality in the units. A great many small tribes which had existed in Caesar's time now ceased to exist, being absorbed into larger ones or grouped into new units. It is possible to give what seems to be a fairly complete survey of the resulting units.

In the south-east, the tribes of Kent were united into a single unit with a capital at Canterbury, Durovernum Cantiacorum. Separated from these by the forest of the Weald was the civitas of the Regni, with its capital at Noviomagus. The district south of the middle Thames became the canton of the Atrebates, with its capital, Calleva, at Silchester. North of the Thames, instead of a tribal unit of Catuvellauni, there was the municipium of Verulam with its territory; and farther east, instead of a civitas of Trinovantes, the colony of Camulodunum. East Anglia became the

168

THE MACHINERY OF GOVERNMENT

canton of the Iceni, with capital Venta Icenorum, the now-deserted site of Caistor-next-Norwich. The Cotswolds, ancient kingdom of the Dobuni, became the civitas Dobunorum, with Corinium (Cirencester) as its capital; beyond this, in the Severn valley, the territory of the colony at Gloucester. South of this, and west of the Atrebates and Regni, was a belt of country which had formed the western fringe of Commius's kingdom; it extended from south-eastern Hampshire obliquely to the Bristol Channel, and was made into an artificial unit under the name of civitas Belgarum, with its capital, Venta Belgarum, at Winchester. Beyond this again lay Dorset, the country of the Durotriges; their capital, Durnovaria, was at Dorchester, close to their prchistoric fortress of Maiden Castle. And the whole western peninsula, probably beginning at the river Axe, was the territory of the Dumnonii, whose capital Isca was on the site of Exeter. West of the Severn were the Silures, whose capital was yet another Venta, now Caerwent. In the midlands, Ratae (Leicester) was the capital of the Coritani; Viroconium (Wrox-eter) that of the Cornovii. Lincolnshire must have been attributed to the colonia of Lincoln. In Wales we have no trace of civitates except for that of the Silures; the other tribes of which we know—Demetae in the south-west, Ordovices in the centre and north-west, Degeangli in the north-east—do not seem to have had self-governing institutions and certainly had no roman-ized capitals. Their land may have been altogether under military government; the Degeangli, for example, were probably administered directly by the legionary commandant at Chester, whose mines and tile-works were in their territory. Finally, the Brigantes formed a civitas with capital (Isurium) at Aldborough. But much of Yorkshire must have been attributed to the colony at York; and the northern Pennines and frontier district as a whole, though historically Brigantian, were doubtless under military administration.

As to the boundaries of these districts we can only guess; and therefore it is difficult to be sure that the list is complete. There is one obvious omission. Whatever may have been the exact status of London, it was so large and so important a town that it must have possessed, not only ordo and magistrates of the standard kind, but sufficient public resources to make it in effect a power in the land. But it was not a tribal capital; it was not originally, and probably never became, a colonia; and there is

THE PROVINCIAL COUNCIL

169

no evidence that it ever became even a municipium. Its wealth was derived from commerce and industry, and in the absence of municipal rank it cannot have possessed the territorium which was peculiar to that rank. The omission of London from the above list, therefore, is more apparent than real; for nothing that we know of London entitles us to regard it as the centre or capital of any district, belonging in a special sense to itself.

It was the policy of the empire to foster self-government on the part of provincials not only in their various tribal units, but to a certain extent even in the province as a whole. The machincry by which this was done depended on the institution of emperor-worship. In each province there was a centre where this worship was carried on by officials specially appointed on behalf of the province as a whole. Representatives from every part of the province (in the case of Britain, no doubt from the tribal civitates) met annually in the concilium provinciae, the provincial council, whose chief business was to elect a provincial high priest for the due performance of the rites of worship, and to raise the money which that performance required. But this council had other duties and powers as well as its primary religious function. It had direct access to the emperor as representing the corporate voice of his subjects in the province, and in that capacity could express its opinion of the merits or defects of a governor. Normally it exercised this right only by passing a vote of thanks to a retiring governor, but on occasion it might pass a vote of censure. No doubt this was rarely done, but cases of it are on record.

In Britain, the meeting-place of the provincial council was the temple of Claudius at Colchester, which was obviously marked out by its history under Cunobelinus to serve as the centre of provincial life and opinion. But the system made a bad beginning. Britain was not a unity, and the attempt to make it into one by setting up a provincial council was premature. This is shown by the revolt of Boudicca, in which the hatred of the rebels was especially directed against the temple of Claudius and all that it implied, that is to say, against the very institution which the Romans had intended to serve as an expression of their subjects' wishes. After this we hear no more of it. That the provincial council continued to exist we cannot doubt. But we cannot even be sure that it continued to meet in Colchester. A fragmentary inscription from London suggests

170

THE MACHINERY OF GOVERNMENT

the possibility that, by the early part of the second century, the

seat of emperor-worship may have been transferred to a place

less dangerously associated with ancient tribal enmities.1 There

is also in London a tombstone to the wife of a servus provincialis,

a slave belonging to the provincial council, which is additional

evidence of the same transference.2

If the question is asked: where was the capital of Roman

Britain? no answer can be given unless we distinguish between

four possible senses of the word capital: religious, civil, finan-

cial, and military. We have seen that the religious capital was

first Colchester, and later London; the move having been

accomplished early in the second century, perhaps immediately

after the Boudiccan revolt. The civil capital would be the seat

of the governor's residence and offices. We do not know where

this was at any time; but what Tacitus says about London

suggests that it was not there in the days of Boudicca. Perhaps

that, too, was first at Colchester and later moved to London,

which would certainly be the most convenient place for it when

London had grown to the largest town in Britain, and the

most accessible. The financial headquarters, as we have seen

(p. 164), were in London. Reference has been made in an

earlier chapter to the tombstone of Julius Classicianus, who

became procurator during the course of Boudicca's rebellion.

This tombstone was found in London, and if, as is probable, it

indicates that Classicianus died before he laid down his office,

the natural inference is that he lived and worked in London,

and that the procuratorial offices were already there just after

the rebellion, perhaps before it also.

Perhaps the likeliest suggestion would be that, before the

Boudiccan revolt, Colchester was the civil and political capital,

London the financial; and that in consequence of the revolt

London became capital in all three senses.

Certainly London was marked out by nature to serve such a

purpose. But it does not follow that the most convenient centre

for the military administration would be found there. London

is between 150 and 200 miles away from each of the three legion-

ary fortresses, and 300 from the Wall, where from Hadrian's

time onwards the military organization of the province had its

1 C.I.L. vii. 22; the sense of the inscription seems to have been 'to the divinity

of the Emperor and to the Province of Britain'; that rendering, and the date given

in the text, are Haverfield's (J.R.S. i. 151).

2 C.I.L. vii. 28.

centre of gravity. By the time of Severus there was at York

something called a domus palatina, an imperial residence, and

although this does not imply that York was in every sense the

capital of Britain, it may not inconceivably point to the exis-

tence there, perhaps from an earlier date than this, of a general

head-quarters for the army. More probably, however, this

domus was merely the residence of the governor of Severus's

new province, Lower Britain.

Such was the general plan of governmental machinery by

the time Roman Britain had crystallized into definitive shape

under Hadrian. The drastic revision of the entire scheme by

Diocletian will be described in a later chapter; here a few minor

alterations, falling between Hadrian's time and the end of the

third century, must be mentioned.

Something has already been said of the Brigantian revolt in

the reign of Antoninus Pius. Pausanias, our authority for it,

says that it was followed by Roman annexation of a great part

of the Brigantian territory. This seems at first sight strange,

since the whole of Brigantia was already well within the Roman

frontier. But it has already been pointed out that one large part

of that region must have belonged, partly as colonists' allot-

ments and partly as attributed land, to the colonia of York, and

that another large part was probably administered not from the

tribal centre at Aldborough, but by the military authorities of

the frontier district. It is therefore easy to understand the words

of Pausanias as meaning that the reduced portion of Brigantia

which in 154 was still governed by the tribal authorities was

now still further reduced by transferring a large part of it to one

or other category of land under direct Roman rule.

Another alternative is even likelier. The documents attesting

the colonial status of York, unlike those for the three other

colonies in Britain, are all of a late date, and are consistent with

the possibility that the colonia there, as distinct from the legion-

ary fortress and the usual canabae forming a settlement outside

its walls, did not begin to exist until the Antonine period. Know-

ing as we do that the Brigantes were punished for their revolt by

'cutting off', as Pausanias puts it, a great part of their land, it

is very probable that what Julius Verus actually did was to

create the colony at York, to give the colonists allotments in the

fertile vale of the Ouse, and to annex to their community in the

form of ager attributus a considerable proportion of the rest. This

272

THE MACHINERY OF GOVERNMENT

is all the more likely because Aldborough, whose walls are

thought to be of Hadrianic date, lies so close to York that if the

colonia there existed in Hadrian’s time it is hard to see how the

allotments and attributed land of the colony could avoid jostling

the Brigantian capital, in which case one might have expected

Hadrian to move the native town to another and more distant

site.

The only other innovation that need be considered here is the

division of Britain by Severus into two provinces. The fact is

baldly stated by Herodian1 as having happened immediately

after his victory over Albinus at Lyons. No hint is given of his

motive in making the division, nor of how it was made; but a

passage of Dio,2 whose evidence is borne out by various inscrip-

tions, tells us that the Second and Twentieth legions were in

upper Britain, the Sixth in lower. An inscription found some

years ago at Bordeaux adds the information that Lincoln, like

York, was in lower Britain.3 From a couple of inscriptions once

existing at Greta Bridge, on the Stainmore road from York to

Carlisle, which mention officials belonging to the upper pro-

vince, no inference can be drawn, except that persons officially

attached to the upper might on occasion be seconded for duty

in the lower.

The general purport of the division, from a military point of

view, is clear, especially when it is recollected that Severus

restored the Second legion to its full strength, recalling its drafts

from the north, and reoccupied in force at least one castellum in

Wales, namely, Segontium at Carnarvon. Whether because he

did not trust the Welsh tribes, or because there was already

reason to fear raids from Ireland (both are possible, but there

is no positive evidence for either), Severus used the retrench-

ment of forces in the north, due to the reorganization of the

frontier on Hadrian’s line, for strengthening his position in the

west; and this new arrangement was accompanied by a new

distinction between a western command, including Wales with

the legionary fortresses naturally belonging to it, and a northern

command comprising Brigantia and the Wall.

The natural result of this arrangement would be that upper

Britain, having two legions, would be governed by a legate

1 iii. 8. 2.

2 Dio, lv. 23.

3 J.R.S. xi. 102: an altar set up in 237 by a ‘sevir Augustalis of the colonies of

Eboracum and Lindum in the province of Britannia Inferior’.

THE TWO PROVINCES OF SEVERUS

173

of consular rank, lower Britain, since it had only one, by a praetorian. When a general concentration of troops was needed the consular would take command. A system of this kind, however, would naturally tend to leave traces in the inscriptions, of which it happens that we possess a fair number for the thirty years following the change; not only is nothing of the kind traceable there, but the texts make it improbable that such a system was actually in force. We find, in the early years of the third century, a governor of consular rank recording his presence in Northumberland without any mention of his belonging strictly to the upper province. Possibly, in view of the exceptional importance of the York command, with its enormous number of auxiliary troops, it was allotted to a consular governor in spite of possessing only one legion. In that case, the two commanding officers being of equal rank, it is possible to accept an explanation of Severus's motive for the change which would not hold good if one were under the other's orders. The adventure of Albinus, he may have thought, proved that the entire army of Britain was too large a force to be left safely in the hands of a single governor. It could be divided in two without seriously weakening the frontier, and for political reasons this was accordingly done. It thus falls into line with the general policy of Severus, according to which frontier armies are so divided up that no one governor commands more than two legions.

From a civil point of view, the division probably implied that London and the greater part of the civilized region fell in upper Britain, whose governor would accordingly use London as his capital, while that of the lower province had his residence at York, this being doubtless the domus palatina to which Severus retired after the conclusion of a campaign in Scotland. How the division between the provinces ran we do not know; most probably from the neighbourhood of the Wash or Fens to that of the Mersey; for East Anglia in Roman times, as to-day, was more closely connected with the Thames basin than with Lincolnshire.

XI

THE PEOPLE

However difficult it may be to give a generalized account of the population of Roman Britain, its character, its distribution, its density, and the changes which it underwent during our period, the attempt is worth making, if only as a framework into which the details to be considered in later chapters may be fitted. The difficulty is due to the fact that here, as usual in ancient history, we have no statistics, and without statistics any sketch of the general state of a country is at best shadowy and vague, and all too probably wrong in essential points. The possibility of attempting it at all depends on the accumulated labour of many generations of archaeologists, studying and describing the remains of the period found in various parts of the country. By compiling these results and assembling them into descriptive lists such as Haverfield contributed to many volumes of the Victoria County Histories, and by plotting the fruits of many such compilations on a map, which has been admirably done in the map of Roman Britain published by the Ordnance Survey, it is possible to obtain a general picture of the population in its relation to the land, a picture which, so long as its limitations, due to the character of the materials on which it depends, are never forgotten, can claim a certain approximation to the truth.

The Ordnance map depicts the distribution of the Romano-British population exactly as it depicts, for example, the Roman road system. Here and there it shows a stretch of road a few miles long, ending in the air, not because the compilers thought that the Romans built roads leading from nowhere to nowhere, but because the farther course of these roads has never been exactly discovered. Similarly, it shows the entire highland zone as, apart from the Roman forts which it contains, almost empty of inhabitants. This is not because we believe it to have been uninhabited, but because sites there, other than these forts, which have been proved by objects found in them to have been inhabited during the Roman period are very rare. Again, it shows large areas within the lowland zone, such as Lincolnshire and the Fens, as sparsely inhabited, because when it was compiled the exploration of these areas had been very incomplete;

THE POPULATION-MAP

175

in both cases much progress has been made in the last few years. Hence, in using the map for the purpose of the present chapter, two limitations must be borne in mind.

First: in proportion as any section of the people was un-romanized, the chances of its being adequately represented on the map diminish. This is a consideration of the utmost impor-

tance for any attempt to estimate from the map the total num-bers of the population. If we believed that practically the whole of the people was appreciably romanized we could rely on the map to give us at least a fair sample not only of its distribution, but of its numbers; whereas, if we suppose any given fraction—say one-half—to have been untouched by Roman civilization, we should be obliged to infer that concerning this half the map tells us nothing whatever.

Secondly, a special density or rarity of entries on one part of the map may be due, either to special density or rarity of the population (that portion of it, of course, with which the map is concerned) in the corresponding part of the country, or else to special intensity or backwardness of archaeological explora-tion in that district. For this reason the map cannot be used for our present purpose apart from some acquaintance with the history of research in different parts of the country.

With these qualifications, the map gives us a scheme of distri-bution somewhat as follows.

In the lowland zone, the areas of primary settlement contain the vast majority of the inhabitants. They are dense in its central ganglion, the chalk plateau of Salisbury Plain; dense on the plateaux of Berkshire and Hampshire and on the chalk ridges that run from these centres into Kent, Sussex, and Dor-set; dense again, though less so, on the ridge that runs by way of the Chilterns and Cambridgeshire to the Wash. The same is conspicuously true of the oolite Cotswold-Northamptonshire-Lincolnshire ridge. But the chief river-valleys, and some low-lying lands, too, have a large population, especially the middle and upper Thames from Staines to Oxford and above, the Cam near Cambridge, the drier parts of the Fens, and the coastal strip on the Sussex-Hampshire borders from Brighton to South-

ampton. The forest of the Weald is conspicuously empty; so, to a less extent, are the heavy lands of Norfolk and Suffolk; so, too, the clay belt between the oolite ridge and the chalk.

Beyond the oolite ridge we enter a region almost devoid of

176

THE PEOPLE

inhabitants. The Severn valley above Gloucester, the whole midland belt stretching thence to Leicester and Derby, and the marshes of the Ouse basin to the north, all these show hardly anything except a few settlements on the Avon, a few posting-stations on the main roads, a town and a group of villas at Leicester. The triangle whose corners lie at Gloucester, Lincoln, and Chester is the most sparsely inhabited region in Britain.

So far, in dealing with the lowland zone, we need make comparatively little allowance for the first source of error. All over that area sites even of the humblest villages so constantly yield traces of Roman influence to the excavator's spade, even if only a few potsherds, that we shall probably not be far out if we assume a certain degree of romanization in every section of the people, down to the humblest: not enough to justify us in calling them romanized in manners and life, but enough to justify us in thinking that their settlements have not been omitted from the map in very large numbers for lack of evidence to date them.

In the highland zone all is changed. Here the degree of romanization is so low that when settlements are dug whose general character would suggest ascribing them to this period, it is not always that they contain even a potsherd to make the dating certain. Although, therefore, the highland zone contained no lack of inhabitants, their unromanized condition has banished them from the map, and their distribution must be gathered from other evidences.

Beyond the belt of sparse population there are, on the fringe of the highland zone, three outposts, as it were, of romanized life. The first is in Monmouthshire and Herefordshire, with a few towns and a sprinkling of villas. The second is in Shropshire, with a large town at Wroxeter and a few villas. The third is in central Yorkshire, the richest of the three.

There is evidence for a dense average population of the highland zone in the character of the Roman towns bordering upon it. The upper Severn basin, the country of the Cornovii, contains very few villas and not many other inhabited sites dated to the Roman period. If we knew nothing of its capital, Wroxeter, we should think of it as a somewhat sparsely inhabited region. But Wroxeter, with 170 acres within its walls, is the fourth town of Britain in point of size. Its public buildings were among the largest and most magnificent in the country. Such

POPULATION: DISTRIBUTION AND DENSITY

a town cannot have existed in a thinly populated district. Though other evidence for a dense population hereabouts is wanting, Wroxeter by itself is evidence enough. Caerwent, again, though a small town in size, is remarkable for the closeness with which houses are packed into its walls; as we shall see, it must have contained at least as many people as Silchester, though its area is smaller by more than half. This points, by the same reasoning, to a dense population in the canton of the Silures. If the Silures and Cornovii were so numerous, we may fairly suspect that the tribes of the highland zone in general were hardly less so, as, indeed, their hill-top towns prove.

As to the distribution of these inhabitants, the highest and wildest parts of the highland zone were hardly inhabited at all: nobody lived in the heart of the Lake District, for example, or on the main plateau of the Pennines. Again, the valley-bottoms, where the heavy mountain rains caused frequent floods, and where woods and swamps made cultivation difficult, were seldom occupied. But on the lower hills, on the knees of the mountains, and on the shelves of drier ground between the hill and the valley, especially when the soil was permeable and not too much soured and sodden with peat, there was a large population, often larger than the same kind of land carries to-day; probably not less in average density than the population of the lowland zone.

In all parts of the country the distribution of the inhabitants during the Roman period has a character which separates it from that of the modern population and assimilates it to the prehistoric. Everywhere it is the areas of primary settlement which were occupied: in the highland zone, the naturally drained slopes above valley-level; in the lowland, the chalks, oolites, gravel, and brick-earth. To-day these areas have everywhere been, if not deserted, at any rate robbed of their primacy by a gradual process through which areas of secondary settlement, heavier lands impossible to cultivate without a great expense of labour in forest-clearance and drainage, have been brought under cultivation. In Wiltshire, the river-valleys that dissect the plateau of Salisbury Plain are to-day cultivated from end to end, and studded with chains of villages; in Roman times they were empty, and the people lived on the plateau itself. The same contrast can be seen in the Berkshire Downs and the vale of the White Horse below them, or the South Downs near

178

THE PEOPLE

Brighton· and the strip of low country along the coast, or the Cotswold plateau and the Worcestershire plain.

The shift of population from the primary areas to the secondary has been, perhaps, the greatest change that Britain has undergone since first it began to be inhabited. It has been described as a 'valleyward' shift, and to some extent the epithet is just; but strictly it was a 'forestward' shift. Its essence was a change of habits by which, instead of choosing land to cultivate where cultivation was easiest, because the land was unencumbered either by dense timber or by excess of moisture, men began to cultivate, undeterred by these obstacles, wherever the land was good. Before the shift took place, wooded and water-logged land was regarded as uninhabitable. After it had begun, such land was reckoned as in no way disqualified for occupation. The result, at lowest, was a great increase in potentially cultivable land and consequently in the total population and wealth of the country; at best it might be an even greater improvement, if the new land brought into cultivation was intrinsically, once it had been cleared and drained, better and more fertile than the old; and in such cases there would be a tendency for the new land not only to rival the old in population, but actually to drain it of inhabitants.

This shift did not take place all at once. But, broadly speaking, in the time of Julius Caesar it had hardly begun; by the time of Domesday it was, if not complete, at any rate far advanced. We think of the Romans as a great civilizing power, and it seems natural to assume that a movement of this kind, resulting in so great a development of the country's unexplored wealth, might have been due to them. But that assumption only seems natural because we unconsciously liken the Roman settlement of Britain to, for example, the English exploration and development of central Africa, which was from the first deliberately planned as a means of increasing the production of wealth by applying European methods of cultivation to soils hitherto, by European standards, neglected.1 How false the analogy is may be seen from a study of the actual distribution of inhabitants in the Roman period, which reveals the fact that,

1 The motive, combined with that of introducing the blessings of the Christian religion, appears, for example, on almost every page of Livingstone's books: Expedition to the Zambesi, 1858–1864, p. 588: 'the favourable soil and climate render it probable that with skill in cultivation this country might be made to exceed many others' [in production of cotton].

PRINCIPLES OF DISTRIBUTION

179

whenever the great forestward shift of population took place, it

did not take place then. Throughout that period the vast mass

of the people, in highland and lowland zones alike, continued

to live as it had lived in prehistoric times on the areas of primary

settlement, nibbling here and there at the forests and planting

a few farms in clearings, but for the most part preserving its

prehistoric habits with regard to choice of cultivable land un-

changed.1

The distribution of inhabitants in the Romano-British coun-

try-side, then, is qualitatively prehistoric, being based for the

overwhelmingly greater part on the areas of primary settlement:

and with regard to the exceptions, with their evidence for a

slight and hesitating valleyward and forestward tendency, it

must be remembered that this tendency had set in before the

Roman conquest, owing to the settlement of the Belgae. Apart

from these exceptions, of which the most remarkable is the

Roman drainage and colonization in the Fenland, the chief

characteristic of the Romano-British distribution is quantitative.

'In Romano-British times practically the whole of Salisbury

Plain, Cranborne Chase, and the Dorset uplands were under

plough'; and a recent survey has shown that the same is, broadly

speaking, true of the South Downs.2 Without changing their

habits, Roman influence apparently stimulated the inhabitants

of Britain to exploit more completely those soils which they

already knew how to till, a fact which probably indicates a

rise of population, as well as the existence of an incentive to

increase production in order to meet the demands of the tax-

collector.

When the attempt is made to estimate not merely the relative

density of this population in various parts of the country, but its

absolute density, the difficulties thicken. We probably know

all the towns of any size that Roman Britain contained, and the

areas of nearly all; excavation has shown us roughly how far

these areas were ever full of houses; consequently we can make

1 The general facts have been laid down once for all by Sir Cyril Fox, whose

intensive study of the Cambridge region proved that a widespread, though sparse,

occupation of forest-land set in during the Roman period (Archaeology of the Cam-

bridge Region, pp. 224-5). But the total results were small, and the Domesday

distribution pattern was the product of a decisive change 'wrought between 650

and 1050 A.D.' (Personality of Britain, ed. 2 (1933), p. 71).

2 O. G. S. Crawford and A. Keiller, Wessex from the Air, p. 9; Antiquity, ix (1935),

p. 443.

180

THE PEOPLE

a fair guess at the total number of town-dwellers. Of villas there were certainly a great number which are still undiscovered; but it is most improbable that the unknown ones were more numerous than the known; and here again the size of a great many has been ascertained, and a guess can be made as to the number of persons they would probably contain. The villages of peasants enjoying a low degree of romanization are known to the number of some 700, and here again we know the size of a good many; but the number of unknown ones is far harder to guess. And finally, when we come to the population of the highland zone, data fail us altogether. The result is, roughly speaking, that we can estimate the more or less romanized inhabitants of Britain, including the army and its dependents, at about half a million souls; but this leaves it an open question what fraction of the whole was sufficiently unromanized to slip through the meshes of our estimate. Perhaps, if we allow another half-million for that, and estimate the total at a round million, we shall not be very far wrong.

It still remains to ask whether during the Roman period this population, taken either as a whole or in its various sections, rose or fell. Much has been written of decline in the population of the Roman empire, but the best modern opinion holds that this decline was probably limited to Greece and Italy, and that in some other regions there was actually an increase.1 We must therefore approach the question, as it affects Britain, with an open mind.

In the case of town-dwellers it is certain that, after rising to a peak in the second century, numbers must have greatly diminished. The evidence for this decline will be discussed in the next chapter. The inhabitants of villas, on the contrary, increased, though the increase here can hardly be thought to balance the decrease in the towns. With regard to the village-dwellers, who in any case made up the large majority of the population, it is hard to say anything definite. The abolition of tribal warfare no doubt at first brought about a rise; but in villages, as in villas, there is evidence that the Roman peace brought with it the practice of infanticide, which would tend to stabilize numbers. Here and there villages seem to have grown larger as time went on; but on the other side of the account we must place some which were deserted. On the whole, it is probable that the rural

1 Rostovtzeff, Social and Economic History of the Roman Empire (1926), p. 328.

POSSIBLE RISE OF POPULATION

181

population tended to rise, rather than fall, under Roman govern-

ment; but the evidence does not allow certainty.

The tendency to take up land in forest-clearings certainly

points to a rise; but we do not know at what time this tendency

was chiefly operative. If we could generalize from the case of

a villa in a forest-clearing in Wychwood, near Oxford, which

is known to have begun its existence in or even before the

Flavian age, we might be disposed to think the tendency charac-

teristic of an early period, when the newly established pax

Romana was causing population to rise, and when equilibrium

had not yet been reimposed.1

The population of the highland zone probably rose a good

deal during the Roman period. That is the natural inference

from the numerous hill-forts that were built both in Wales and

in the lowlands. Probably the relative or absolute cessation of

tribal war caused an increase which was not checked, as it was

in the lowland zone, by various by-products of civilization.

In an earlier chapter some account has been given of the

physical characteristics of the Britons before the Roman in-

vasion. The question must now be asked how far these charac-

teristics were affected by the invasion and the events to which

it led. There are two classes of evidence to be considered:

historical evidence for immigration, and anthropological evi-

dence for change in physical type.

There was certainly a great deal of immigration, especially

but not exclusively at the beginning of our period. The Roman

conquest itself brought into Britain some 40,000 soldiers of non-

British origin; when to this number we add traders and camp-

followers of every kind we cannot put the number of immigrants

at less than 100,000, and the probability is that it was even

larger. And during the generation that followed the landing of

Claudius's army the influx of foreigners must have continued,

though it probably declined and, by the end of the first century,

was no longer very considerable. The great majority of these

early settlers either made up the army, and therefore went where

the legions went, or else were engaged in trade and may be

supposed to have taken up their abode for the most part in

1

Antiquity, iii (1929), pp. 261–76. In reconsidering the views there stated, I

have thought it desirable to allow a much larger figure for unromanized village-

dwellers, and to attach more weight to the possibility of a rise in rural population

during the Roman period. I have to thank Mr. H. J. Randall and Dr. R. E. M.

Wheeler for observations on the former point, ibid. iv (1930), pp. 80–95.

3720.1

N

182

THE PEOPLE

London, and to a lesser extent in Colchester, Verulam, Chichester, Silchester, and the other Roman or romanized towns of the south-east.

If we ask where they came from, the answer, so far as we can give it, is derived from a study of early tombstones. We have about fifty tombstones of the first or very early second century which give some account of the origin and occupation of such settlers. Of these, forty-five commemorate legionaries; the rest were either civil officials or traders. From Italy proper there are only four, two of them soldiers; from Cisalpine Gaul nine, all soldiers; from the Province in southern Gaul four, all soldiers; from the rest of Gaul eleven, seven of them soldiers; from Spain six soldiers; from Germany a single soldier; from the Danubian provinces fourteen soldiers; from Greece, Asia, and Syria five, two of them soldiers. On the evidence of these tombstones, the legions of the invasion would seem to have been recruited chiefly in Gaul, whether Cisalpine, Provincial, or Lugdunese, in the Danubian provinces, and in Spain, whereas the traders and officials apparently came from Italy, northern Gaul, and Greece.

Inscriptions of later date, as we should expect, add comparatively little. The main fact which they reveal is that, after the original invasion, reinforcements for the army, when needed on a large scale, came almost exclusively from Germany and to a less extent from the upper Danubian provinces. The influx of German soldiers, whether individually or (still more) in blocks, is shown to have been very considerable in the late second century and early third, after which inscriptions fail us. In general, units of the Roman army, from whatever source they were originally derived, tended to settle down and take root in the country where they were quartered. Among the auxiliary regiments in the army of Britain, over half came from northern Gaul and the Rhineland, a quarter from the Danubian provinces, and almost all the rest from Spain. The great majority were thus derived from regions where the people differed little either in physical character or in civilization from the Britons; and when once they had settled in Britain they would recruit themselves, according to Roman custom, on the spot, and consequently would assimilate themselves somewhat rapidly to the country of their adoption. The evidence of inscriptions certainly tends to show that this local recruitment proved inadequate for the maintenance of their strength; but it also shows that when

IMMIGRATION FROM THE CONTINENT

183

new drafts were needed from elsewhere they were drawn mostly

from the Rhineland.

Thus, apart from the initial influx of legionaries, mostly from

Gaul, the Danube, and Spain, the physical influence of the

Roman army on the population of Britain, as judged by inscrip-

tions, cannot have been very great. We should expect it to

show, if at all, in a tendency towards the establishment of Ger-

man physical characteristics, and we should expect this tendency

to appear only in the north.

There is another source of racial mixture which must be con-

sidered. Especially during the later period of the empire, it was

not unusual to transplant considerable masses of inhabitants

from one district to another. Burgundians and Vandals from

northern Germany were planted in Britain in this way by Probus

in the third century, Alamanni from the upper Rhine by Valen-

tinian I in the fourth.1

Apart from these transplantations, the historical evidence

applies almost exclusively to the earliest period of the occupa-

tion. That of physical anthropology applies only to its later

phases; for it depends on the measurements of bones, especially

skulls, and this becomes available only in the third century,

when cremation went out of fashion and was replaced by in-

humation. Thus the historical and anthropological evidence

are mutually complementary: the first gives an account of the

immigration and settlement which occurred chiefly in the first

generation of our period; the second gives us a view of the

population some two to three centuries later, when, in spite of

easy communications both within Britain and between it and

the rest of the empire, it is reasonable to assume that a stable

condition had been reached.

The results of anthropological study are curiously uniform

and definite.2 The Romano-British physical type, whose general

features have already been described in an earlier chapter (supra,

ch. ii, pp. 16–18), occurs all over Britain in skeletons of the third

and fourth centuries. It is not confined to the native villages;

skeletons found in villas and in the cemeteries of towns show

exactly the same type. The importance of this fact appears

when it is recollected that the native Romans or central Italians

1 Zosimus, i. 66; Ammianus Marcellinus, xxix. 4.

2 Buxton, ‘Racial Affinities of the Romano-Britons’, in J.R.S. xxv (1933),

pp. 35–54.

184

THE PEOPLE

of the time belonged to an altogether different type, being as definitely round-headed as the Britons were long-headed; while other inhabitants of the Mediterranean region such as Maltese, Sicilians, and ancient inhabitants of Egypt and Spain, though they resembled the Britons in being long-headed, differed sharply from them in the very much smaller size of their skulls. Thus the absolute measurements of the average British skull in length and breadth differentiate it clearly from one belonging to the so-called Mediterranean race, while the proportion between these measurements distinguishes it no less clearly from one of the Alpine race, to which the Romans belonged.

When allowance is made for normal variation from the average, skulls of the Roman period found in Britain belong with surprising uniformity to the standard Romano-British type. Even in the legionary cemetery at York, from which we possess a large number of bones, there is no trace of the Alpine type to which a large number of the Claudian legionaries must have belonged, and which is admirably represented by the portrait-tombstone of the centurion Favonius Facilis at Colchester. In other words, by the time cremation gave place to inhumation, the effect of that Alpine immigration on the physique even of the legions themselves had disappeared, and even the army was no less British in physical type than the ordinary townsfolk or villa-dwellers or peasants. Or rather, no less Nordic, if the word is understood in its proper technical sense, as a label for the tall and long-headed race which in prehistoric times inhabited the whole of north-western Europe, though on the Continent that race is now confined to the lands bordering on the Baltic and North Seas. For within that race it is possible to recognize varieties differing slightly from one another; and there is a difference of this kind between the standard types of Romano-Britons and Anglo-Saxons. The legionary cemetery at York has yielded a large number of male skulls which, though not very different from Romano-British, are in size and shape identical with Anglo-Saxon. They are, no doubt, to be explained by reference to the reinforcements sent at various times, as we have already seen, to the British frontier armies from the Rhineland.

Groups of skulls have turned up here and there in England which closely resemble the Italian type. Unfortunately, not one of these groups can be dated by archaeological evidence to the Roman period; but as one was found at Richborough there is a

TRACES OF ITALIAN AND GERMANIC BLOOD

185

possibility that it may represent a population which was either

descended with little native admixture from original Italian

settlers, or, more likcly, onc which was reïnforced from time to

time by new-comers from the same source. Another was found

at Dunstable in a barrow of the fifth or sixth century; here, too,

therefore, there is a suggestion of a settlement whose blood was

in the main Italian. A third and very interesting case is that of

a large collection of skulls found at Spitalfields in London,

belonging to people who appear to hạve perished in some

pestilence or other catastrophe, and to have been buried in

cart-loads. Their physical character so closely resembles that

of skulls from Pompeii that the evidence for a population of

Italian blood is here very strong; and, as we already know,

London is the place where such a population is most likely to

have existed. It is tempting to see here the victims of Boudicca's

massacre; for though at that date cremation was in general

use it may have been dispensed with in clearing up the ruins

of a destroyed city, where all the timber that could be had

was needed for reconstruction.

The inference from this anthropological material is definite. In

some places, notably Richborough and London, there may have

been populations whose debt to an Italian ancestry was visible

in their bodily appearance; in the north, a Germanic physical

type may have had its influence on the army; but elsewhere,

in town and country alike, the later part of the Roman period

saw Britain inhabited by a race highly uniform in physique,

and in that respect definitely and characteristically British,

owing nothing physically to the Italian influence to which,

culturally, they owed so much.

XII

THE TOWNS

The Mediterranean world is a world of town-dwellers. By her early tradition, Rome, like the cities of Greece, rested upon an economic foundation of landed property, and in the last resort implied the conception of the citizen as farmer; but politically and socially. the civilization of the Mediterranean depended on the life of the town, and implied the conception of the farmer as citizen. And citizenship was no metaphorical expression or legal fiction; it meant bodily presence in the city, personal attendance at assembly and law-court, participation in a kind of life which could not be lived except in towns. Deep in the mind of every Roman, as in the mind of every Greek, was the unquestioned conviction which Aristotle put into words: that what raised man above the level of barbarism, in which he was a merely economic being, and enabled him to develop the higher faculties which in the barbarian are only latent, to live well instead of merely living, was his membership of an actual, physical city. Man's bodily and animal existence might be satisfied by the country; his spiritual needs could only be satisfied by the town; hence the town was at once the symptom and the symbol of all that was highest and most precious in human life, all that raises man above the beasts of the field.

It is necessary to dwell on this conception, because to us it is a little strange. We inherit the words that once expressed it: words like politics from the Greek and words like civility and civilization and citizenship from the Roman; but for us these words are sublimated into metaphor, and we have forgotten, what the Greek and Roman never forgot, that originally they had no meaning in abstraction from the bricks and stones of the city itself. This is because we are heirs of another tradition. Historians of art have shown us that the northern European peoples have never wholly accepted the principles of Mediterranean art, according to which the artist's highest aim is faithfully to depict the human body; and that, with them, the artist is most himself when he is designing formal patterns. There is a like difference between their conceptions of civilization. For the northern peoples, public life needs no town. Its elements already exist in every man's household; and its higher forms

CIVILIZATION AND URBANIZATION

187

crystallize into shape round a tree or a stone where men meet together, and culminate in the court of a king who lives nomadi-cally from manor to manor.

The Greek expressed his yearning after that good life, without which mere life is worthless, by speaking of his city as divine, and of himself, in the words of Pericles, as in love with her. For men of the north, these same yearnings find expression rather in the words of Gunnar, when he turned back from exile and safety to the meadows and homestead of his farm and the death foretold him. ‘Fair is the Lithe; so fair that it has never seemed to me so fair; the corn fields are white to harvest, and the home mead is mown; and now I will ride back home, and not fare abroad at all.’1 And northern men, in spite of all the south has taught them, still tend to think rather in Gunnar’s way than in that of Pericles: conceiving the town as a merely economic fact, a place where man only makes a livelihood, and finding room for the development of his higher faculties only in the country.

In northern countries to-day these two heritages are so inter-twined that neither by itself can satisfy all man’s desires. But when the Romans conquered Britain the two were still separate, and that conquest, in one of its most significant aspects, is the story of their first encounter. Julius Caesar grasped and stated once for all the essential character of contemporary British life, as it appeared by contrast with that of his own country, when he said that Britain was rich in people and cattle, and covered with farms, but that all they had for towns were tracts of wood-land fenced against assault.2 Nor is the contrast invalidated by pointing to the great hill-top towns of the west like Maiden Castle, which Caesar did not know, or to the growth of native towns in the south-east after his time. These hut-clusters were in no sense the nuclei and symbols of British civilization; they were not so much cities as slums.

To convert Britain into a province of the Roman empire, then, was to civilize it in the most literal sense of the word: to furnish it with towns. The same problem had already confronted the Romans in northern Gaul, and the solution adopted there was applicable here too. The hill-top fortresses of the Gallic tribes, places like Alesia and Gergovia and Bibracte, stood not only for tribal warfare and for a freedom that was inconsistent with

1 Njáls saga, The Story of Burnt Njal, tr. G. W. Dasent, ch. lxxiv.

2 Putting together Bell. Gall. v. 12, § 3 and 21, § 3.

188

THE TOWNS

imperial unity, but also for a barbarism inconsistent with

romanization in manners and in the whole spirit of public life.

Where they had not been already destroyed in the wars that

ended Gallic independence, the policy of Rome was, in general,

to decree their abandonment, and to build in the plains below

them new cities after the Roman style, properly planned with

straight streets at right angles, dignified with public buildings,

and surrounded by walls that were at once a defence and an

ornament. In such a city the ancient tribe was still to recognize

its political centre, but its institutions underwent a change

parallel to, and symbolized by, the change from hill-fort to

city.

By the time Britain was conquered, this change in Gaul was

half a century old, and had proved its soundness. Naturally,

therefore, it provided a formula for Britain. If the Gaulish

tribes could be converted into self-governing units of the empire

by giving each a romanized capital, the same could be done

here. In some cases we can even trace a process, identical with

that which had created Augustodunum in the plains below the

hill of Bibracte, by which a hill-top town was deserted and a

new city built close by: thus the settlement in Prae Wood

was replaced by Verulam, Maiden Castle by Dorchester, Llan-

melin by Caerwent. In others, the appropriate valley site was

apparently already occupied, as at Silchester or Winchester or

Chichester or Canterbury, and all that was necessary was to

convert the hut-cluster into a romanized city.

But there was a difference between the two cases. Northern

Gaul before Augustus had developed its towns a good deal

further than Britain before Claudius, and to that extent was

a good deal better prepared for the change. The story of the

Roman wars in Britain contains no sieges comparable with

those which Caesar describes in Gaul; the towns of pre-Roman

Britain were for the most part either less populous or less power-

fully fortified than those of pre-Roman Gaul, and most of them

must have fallen at the first assault. Caesar carried Bigbury

and Wheathampstead by a coup de main; Claudius walked into

Camulodunum; Maiden Castle itself shows no trace of Roman

siege-works thrown up across its ditches or against its ramparts.

And inside British towns, the spade has brought to light no such

houses, squarely built and alined on streets, as those of Gaulish

Bibracte. If it is true, as historians have said it is, that the towns

ORIGINS OF ROMANIZED TOWN-LIFE 189

of Roman Gaul survived the barbarian invasions when those of

Britain went under (and although the contrast has been exag-

gerated there is a grain of truth in it), some part of the explana-

tion may lie in the fact that in Britain the idea of living in towns

was a newer thing than it was in Gaul, and that the Roman

town-building movement was in consequence more artificial.

It also proceeded far more slowly. Excavations at Silchester,

Caerwent, and elsewhere have given us a very complete picture

of wha: a Romano-British town was like when it arrived at its

full development; but for an accurate view of the stages by

which :his development was reached we have to rely almost

wholly on work done since 1912, when the Society of Antiquaries

began their work at Wroxeter; and especially on the same

Society's excavations of 1930 and later in Verulam. The impor-

tance of this last piece of work is twofold. Partly it is due to the

modern methods available for the exploration itself, which have

yielded historical data that were lacking in, for example, the

excavation reports on Silchester; partly to the intrinsic impor-

tance of the site, than which no town in Britain could have given

us fuller and more precise evidence for the character of Romano-

British urban life at its best.

So far as excavation has recovered the buildings of the

earliest Verulam, the town which Tacitus calls a municipium

and which Boudicca's men destroyed, they were of a very

simple kind: mere huts, clustered beside Watling Street and

spreading southward in ribbon-development along it, with mud

floors and walls of wattle and daub, not very different from the

huts of pre-Roman days. It is plain that the Roman govern-

ment in the time of Claudius made no attempt to create a town

on the Roman model; and this is all the more interesting in view

of their readiness to make it a municipium. Backwardness in

material civilization was evidently, to the Romans of those days,

no barrier to the enjoyment of the highest legal privileges.

This earliest town developed rapidly. Industries on a small

scale grew up in it. The equipment of daily life was hardly more

romanized than the architecture; of all the pottery found in its

huts, three-quarters is of the Belgic kind used before the con-

quest, which implies that its inhabitants were for the most part

not foreign traders, flocking in to exploit the new province (for

these would have built houses of a far more Roman kind, and

brought household goods such as to furnish a civilized life), but

190

THE TOWNS

Catuellaunian tribesmen, not of the upper classes alone, ready to adapt themselves to the new régime and make a profit out of the new fashions it had introduced. We catch here a glimpse of the process by which the Celtic style in arts and crafts so rapidly died out and gave place to the general reign of Roman fashion. The Belgic tribes had never been deeply imbued with Celtic taste, and the same romanizing movement that had flourished under Cunobelinus was now being carried a stage farther in the workshops of Claudian Verulam.

A similar process was going on elsewhere. It must have been at this time that Cogidubnus began royally to adorn his capital of Noviomagus with a classical temple and doubtless other buildings in the Roman style; the colonia of Colchester was building the great temple of Claudius, a structure more than twice the size, measured in floor-space, of the famous Maison Carrée at Nîmes; London was a vast and swiftly growing collection of military storehouses and the dwellings and warehouses of traders. Other tribal capitals too must have been growing quickly, if somewhat chaotically and squalidly, with sharp contrast between a few handsome public buildings, where the authorities were anxious to show themselves in the van of progress, and a mushroom-growth of hovels where artisans hammered and chiselled and modelled, blew at forges, or handled crucibles and hot metal with the tongs.

So quick and so complete had been the success of the Roman armies, so ready the conquered British to accept the fait accompli, that no one thought of danger. These straggling, kaleidoscopic townships went unwalled, without even a palisade against possible enemies. The tempest of Boudicca's men fell upon them unawares, and wiped them clean away.

For a progressive and thriving town, fire is not an unmixed evil. Verulam rose from its ashes not indeed transfigured, like Wren's London as it ought to have been, but better built with rectangular wooden houses, and fenced foursquare with a ditch and palisaded bank. The men who made these defences had confidence in their city: they enclosed 140 acres of ground, an area only exceeded by three other towns in Britain even at the height of their prosperity, and nearly half the size of London itself. The houses which now begin to appear are the standard smaller town-house of all the north-western provinces. They are detached buildings, long and narrow, one gable-end facing the

street. At first they were generally built in timber, with this

front end open to serve as a shop; the whole house is like a large

cart-shed or garage, where people lived and plied their industry

and sold the produce. These open-fronted wooden houses are

typical of the late first century in Britain; in construction a link

between the ruder huts of an earlier stage and the similar stone

houses of a later; characteristic of a time when towns were

growing fast and buildings had to be put up cheaply and

quickly, but when romanization had already reached the

point of dictating the architecture of even the humblest town-

dwellings.

Such were the houses of the poorer townsfolk all over Britain

in the Flavian age, when Agricola, and doubtless other gover-

nors as well, began officially to encourage town life. By this time

the towns were already serving at least three purposes. Legally

and politically they were seats of local government, where the

magistrates of the tribe did justice and its ordo met to transact

its public business; commercially they were markets for the

country round; and industrially they were the homes of artisans

working busily to supply the ever-growing demand for goods

made in the Roman style. But in architecture and in the

character of their household furnishings they were still very

primitive, far behind the towns of northern Gaul; at Verulam,

for example, a quarter of the pottery used in the Flavian period

was still purely native in fashion. Their social life was British

rather than Roman, and even their magistrates and decurions

were not always at ease in the Latin tongue. This lag of equip-

ment behind function was what Agricola set out to correct. He

urged the tribal senates to adopt town-planning schemes and to

sink money in handsome buildings, temples, markets, halls of

justice, public baths; he encouraged their members to learn

Latin, adopt the Roman style in manners and dress, and build

themselves houses such as Roman gentlemen should live in. If

money was lacking for these public works, he advanced it from

government funds; if advice was needed, he furnished technical

experts.

Thus, by the end of the first century, the Roman towns in

Britain had laid down the main lines of their development.

Everywhere they had chessboard street-plans, enclosing blocks

whose size and shape varied from town to town, but whose

regularity bears the hall-mark of the Roman surveyor. In or

192

THE TOWNS

near the centre of each town was a forum, a market-square surrounded on three sides by shops and flanked on the fourth by a basilica, an aisled and colonnaded hall such as any modern town of like size might envy. In a little tribal capital like Silchester, a bare hundred acres in extent, the basilica measured 240 feet by nearly 60, and is estimated to have stood 60 feet high. To compare this with a modern town hall, one may reflect that, if provided with chairs in the modern style, it would easily have seated 4,000 people: twice as many as, according to the current estimate, the city contained. Public baths were built where hundreds could bathe at once; the Flavian bathhouse at Wroxeter had hot and cold rooms each measuring 80 by 35 feet, and the rest to match. Amphitheatres, with banks of earth carrying wooden seats capable of holding some thousands of spectators, were erected outside the chief towns.

The first thing that strikes the spectator of this development is its scale. Even down to the present day, no English town has ever been so lavishly supplied with public buildings, relatively to the size of its population, as was the average tribal capital by the end of the first century. A town hall that would hold every man, woman, and child in the city, and more; a second building where they might all watch shows at once; another, where every one of them might bathe daily; such things are hard to believe in these times, when all the public halls and theatres and picture-palaces of an English town will together hold only perhaps a tenth of its population. These public buildings were, no doubt, planned and executed by a small autocratic body; but they were not intended for the use of that body; they were patently designed for the use of the entire city populace.

The successive enlargements of the public bath-house at Silchester, for example, show how its accommodation must have been increased for every increase in the population of the town. Just as emperors provided amusement and sanitation for the population of Rome, so, according to their scale, these local governments of provincial districts, each trying to make its own city a miniature Rome, provided the same gifts for all the people within their city boundaries. Free corn they may not have had; the humbler householders worked for their living; but in other respects the formula panem et circenses held good no less in the little towns of Britain than in the capital of the empire. As for the upkeep of these huge public services, that was met out of

PUBLIC BUILDINGS

193

the tribal exchequer, fed by taxes on its members and by voluntary contributions from its rich men.

We can glean a hint or two even as to the technical staff that helped the tribal authorities to plan these buildings. The forum of the average Romano-British town is not designed on the model of the typical forum as it existed in Roman towns of the south. It seems rather to have been planned after the pattern of the head-quarters building in a legionary fortress. The resemblance even extends to measurements: the forum of an ordinary small town is laid out like the central building of a fortress for one legion; that of Wroxeter, where once two legions lay encamped, is like the head-quarters of a double fortress. Again, the private houses in these towns have little resemblance to the ordinary town-house of Italy; they have no atrium and peristyle, like the houses at Pompeii, nor are they tenement-houses like those of Ostia and Rome. The larger ones are courtyard-houses like the commandant's house at a military site; the smaller resemble the canabae commonly found in the civil settlement outside such a fort. In these ways the planning of the Romano-British town betrays the hand of the military architect; the experts whom Agricola lent to the tribal senates must have been his own army engineers.

Assuming that during the Flavian period tribal capitals thus planned were built at a dozen places in Britain, mostly in the south-east, but extending westwards to Exeter, Caerwent, and Wroxeter, and northwards to the Brigantian Aldborough, it is easy to see what one result must have been. Each town was a focus of intense romanizing influence; but this influence operated primarily on its own inhabitants, and only to a very small extent on any one else. Farmers and villagers living at some distance from the town, even in the most favourable cases, must have been very much less affected by it than even the humblest residents in the town itself. These country-dwellers, or at any rate the more substantial of them, could vote in the city as citizens of the tribe at the annual election of magistrates. They would visit the town when they were involved in cases at law. Many of them would go there now and then to market, and see something of town life. But in no case, unless they were rich landowners who were members of the ordo and therefore probably had town-houses of their own, could the romanizing power of the town work upon them as it would work upon its

194

THE TOWNS

permanent inhabitants. The Flavian movement for the development of towns, therefore, would inevitably divide the population

of Britain into two classes, distinct not in social standing or in legal status or in blood, but in habits and education and culture: one somewhat intensely romanized, the other romanized

in a much lower degree or hardly at all. And there is good evidence that this was what actually happened. The townsfolk,

even the poorest and lowest in the social scale, learnt Latin; learnt in many cases to read and write; learnt to live in a Roman

way. The country-folk, and in particular the villagers, acquired hardly anything of all this. The larger landowners, as we shall

see in the next chapter, became to some extent romanized, roughly in proportion to their wealth; the peasants remained

Celtic.

How far this vast plan of town-building was actually carried out during the last quarter of the first century is a question that

cannot be answered in detail. But there is evidence that the sudden change in their habits was not accepted, even by the

tribal aristocracies, without some reluctance. Tacitus, describing the work done by Agricola in promoting it, tells us not only

that he gave ‘private advice and public assistance’, but that he praised those who were forward to accept his policy and re-

buked those who were backward, ‘so that competition for his favour acquired the force of compulsion’.1 This passage shows

not only that the responsibility for building romanized cities rests on the Roman government, as their official policy imposed

upon the British tribes by what in practice amounted to coercion, but also that a certain degree of backwardness to obey that

command was perceptible even to the governor who most skilfully and energetically pursued the policy. The tribal aristo-

cracies must have found that the habits of a people are not changed without difficulty; they probably found, too, that even

with the help of government grants the money needed for these vast buildings was not easy to come by. They evidently felt that,

in view of these obstacles, the policy was one into which they could not throw themselves with much enthusiasm.

This reluctance was no doubt least in the regions nearest to the Continent, where romanization had already taken root,

and increased towards the north and the west. It is the less surprising to find material evidence of it at Wroxeter, on the

1 Agric. xxi, § 1.

DEVELOPMENT UNDER HADRIAN

195

north Welsh borders, in a striking shape. The great public

baths of that town, as laid out in the Flavian period, were never

finished. Whether because interest declined or because money

ran short, the building, which would have been among the

finest in Britain, remained a mere shell of brickwork, its roof

never put on, its walls never plastered, and its furnaces never lit.

There is some reason to think the Wroxeter baths no isolated

case. They seem to be symptomatic of a failure to carry the

Flavian town-building policy to its logical conclusion. Such a

policy would be complete, as recent excavation at various sites

has shown, only when every town was surrounded by walls

comparable in massiveness and in architectural character with

its public buildings. It used to be thought that the towns of

so peaceful a province, sheltered behind its entanglement of

frontier-garrisons, were originally designed to be unfortified,

and that they provided themselves with walls only when the

troublous times of the later imperial age made them necessary.

This is now known to be a mistake. The colonia of Colchester,

undefended at first, was furnished with magnificent walls and

monumental gates soon after the Boudiccan rebellion. Veru-

lam, as we saw, was simultaneously given an earthwork and

palisade defence. At Caerwent and probably at Silchester

earthen defences of the same kind were thrown up at an early

date. The walls of London were certainly built not later than

the first half of the second century; it is now thought that they

date from the reign of Hadrian. And it was not until the reign

of Hadrian that the earthworks of Verulam were replaced by

stone walls.

Hadrian, according to his biography in the Historia Augusta,

'set many things right in Britain', besides building the Wall.

The excavations of Wroxeter and Verulam have given the words

a new meaning. They reveal Hadrian as the emperor who re-

vived the town-building policy of the Flavian age, rescued it

from stagnation, and carried it through to within sight of the

end. At Wroxeter, where the ruin of the half-finished baths

stood as a monument of failure and bankruptcy in the very

centre of the town, he encouraged the authorities of the Cornovii

to demolish it, to build a splendid forum in its place, and to

erect a bath-house on a new site. The story is revealed by the

inscription that stood over the gateway of the forum, happily

preserved where it fell, dating its erection to Hadrian's reign

196

THE TOWNS

and its completion to A.D. 130, some eight years after his visit to Britain. At Verulam, where houses had begun to straggle southward along Watling Street outside the Flavian earthwork, a new effort made in his reign swept away the earlier defences and laid out a fresh line, enclosing 200 acres of ground and carrying a massive wall of flint and brick, whose gates were not unworthy to be compared with those of the great Gaulish cities.

These facts show that when the town-building policy was no longer being constantly urged upon the British tribal senates by a governor like Agricola, it failed to preserve its momentum. The reasons were probably economic. The capital expense involved must always have been disproportionate to the tribal revenue, and could be met only by constant help from the central government or by remorseless taxation. But whatever the reasons may have been, the fact is that the central government which had initiated the policy found itself obliged to keep up a constant pressure if it was to be fully carried out.

Even with Hadrian, the development of the Romano-British towns was not complete. The stone walls of Caerwent seem not to have been built until the second half of the second century; the theatre at Verulam about the middle of the century. It was the age of the Antonines that saw the towns of Britain reach their apogee. This was the time when their population was largest, their public buildings most splendid, and their private houses most luxurious. At this point, therefore, we may pause for a moment to survey their achievement and estimate the return they gave for all the effort that had gone to their creation.

We find tribal capitals to the number of a dozen: Canterbury, Chichester, Winchester, Silchester, Cirencester, Dorchester, and Exeter south of the Thames; north of it Leicester, Wroxeter, Caerwent; and Aldborough beyond the Humber. Verulam, too, was in fact a tribal capital, differing from the rest in promotion to municipal rank. These represent the romanized urban life of the Britons. London, the largest of all, and by now almost certainly the seat of government, was perhaps rather cosmopolitan than British, though doubtless its inhabitants were for the most part natives of the province. Bath, too, a luxurious and fashionable health-resort, was a cosmopolitan town that stood outside the tribal system. The colonies of Colchester, Gloucester, Lincoln, and York, in spite of their military origin, were becoming increasingly British as the British element came

to preponderate in the army, and were growing less and less

different in character from the ordinary run of the larger towns.

Smaller towns were numerous. We know about fifty, varying

from busy industrial centres to mere posting-stations along the

main roads.

In spite of the scale and magnificence of their public build-

ings, so far in excess of our ordinary modern standards, all these

towns, compared with our own, were very small. London con-

tained within its walls 330 acres; Cirencester 240; Verulam 200;

Wroxeter 170; Colchester 108; Silchester only 100. Some of

the towns which in the preceding paragraph have been classified

as large covered no more than 30 or 40 acres. Nor were the

inhabitants as a rule at all closely packed within the defences.

There were no blocks of flats. The houses were detached and

stood with plenty of elbow-room. At Silchester, where the en-

tire town-plan has been recovered, at Caerwent, where the plan

is hardly less complete, and at Verulam, where trenching has

been done with the express purpose of settling the question, it

is proved that there were considerable areas towards the out-

skirts of the town where no houses were ever built. The walls

were evidently laid out on a generous scale to make room for

large increases of population, and more land was included than

was ever required. When we ask ourselves, therefore, what

population a town of a given size probably contained, we must

be guided by the analogy not of a densely packed modern city

with houses touching one another all along its streets, but of a

residential town where they stand free.

In Silchester, with 100 acres of land, there were only 80

houses. Half of these are of the humblest kind, inhabited doubt-

less by a single family with few slaves or none. Even the largest

could hardly contain a very large body of slaves in addition to

the family. It has been estimated1 that Silchester might have

contained as many as 2,000 inhabitants; as this implies an aver-

age of 25 to a house, it ought to be considered extremely gener-

ous, and any one who proposed to reduce it by half would be

difficult to answer. The evidence of Verulam shows that Sil-

chester was not peculiar in the smallness of even its maximum

population relatively to its size.

1 Oman, England before the Norman Conquest, p. 181. In view of the acknowledgement in the preface, we may assume this estimate to carry the authority of Haverfield's approval.

3720.1

198

THE TOWNS

But there was no constant ratio of population to area. Caerwent was less than half the size of Silchester, but it seems to have contained quite as many houses, if not more: judging by the way they stand in the excavated insulae there may well have been a hundred in the 44 acres enclosed by the walls. In this Caerwent was perhaps exceptional. It seems to have been laid out as a city almost simultaneously with the neighbouring legionary fortress at Caerleon; in its rectangular plan and in its area it owes something to the model of such a fortress; unlike many Romano-British towns, it had defences from the first, owing, no doubt, to the wildness of the hardly conquered frontier district in which it was planted; and the later growth of population had to accommodate itself as best it could to this pre-existing plan. At Silchester, on the contrary, the defences, even in their earlier form as earthworks, seem designed so as to accommodate themselves to a pre-existing town with streets laid down and houses already spreading along them. Caistor-next-Norwich in this respect resembles Caerwent; Cirencester and Wroxeter and London are like Verulam or Silchester.

We may therefore assume an average population of one or two thousand in the smaller tribal capitals, rising to two or three times that number in a place like Wroxeter and even more at Cirencester or Verulam. But even Verulam in the Antonine age is likely to have held, perhaps, nearer 5,000 inhabitants than 10,000; London itself possibly not more than 15,000.1

The smallness of these town populations makes it easier to conceive their relation to the life of the country as a whole. Economically, the towns were parasitic on the country-side. They had to be fed by it, and the goods they produced, together with the services they rendered as markets and trading-centres, were no adequate return for the food they consumed and the expenditure which they demanded for the upkeep of their public services. They had their industries; but these consisted only to a small extent in the production of goods needed in the country; most of them were luxury-trades whose produce was mostly used in the towns themselves. They did a large business in retail trade, selling pottery made in Gaul and other imports,

1 The discrepancy between these conjectural figures and the statement of Tacitus that Boudicca massacred 70,000 persons at Colchester, London, and Verulam is not necessarily fatal to the conjecture. The estimate for Boudicca's victims is likely to be much exaggerated; it must also include refugees.

THEIR GENERAL CHARACTER

199

but, here again, the total quantity of these goods which found

its way into the country districts was the barest fraction of what

the towns consumed. From the strictly economic point of view

the towns were a luxury. Their function was cultural and politi-

cal. They stood for the decencies and elegances of civilized life,

and they provided a link between the Roman government and

the mass of the people, to whom those decencies and elegances

were things out of reach. Their populations, rich and poor alike,

thus formed a privileged section of the people, privileged to

enjoy the blessings of romanization at the expense of the country-

folk.

The strictly economic point of view was not, of course, that

which the central government had in mind when it forced this

policy on the Britons. Its motives, so far as they were conscious

at all, were partly political, the desire to find means of govern-

ing a conquered people, and partly cultural, the desire to create

wherever possible that urban life which to the Roman mind was

identical with the best life. These motives so completely domi-

nated Roman policy that it is hardly relevant to judge its results

by economic standards, to insist that town life in Britain was a

luxury, and to ask whether it was a luxury that Britain could

afford. But if we do insist upon raising this question, the answer

to it will depend on the ability of Britain to pay for the luxury,

and this again will depend on the outlay which it required.

It is the very smallness of the towns that justifies in a strictly

economic sense, if anything can, the policy of creating them.

The general aspect of such a town in the Antonine period is

familiar from excavations. In the open centre of the forum

markets were held, while in the shops opening inwards upon the

market-place, and sometimes outwards too upon the surround-

ing streets, retail trade was carried on. The excavations at

Wroxeter have shown how one of these shops might be full of

Samian pottery from Gaul; another might specialize in coarser

pottery; in another, a packing-case full of whetstones awaited

the customer. In the basilica the public business of the tribe

was being done, and behind it was a row of offices where the

magistrates worked in private with their clerical and technical

staff. In the streets, which were well paved and well drained,

and provided with side-walks for pedestrians, the older wooden

shacks had been replaced by well-built houses, half-timbered

above and roofed with tiles and slates, but standing on stone

200

THE TOWNS

foundations. In the smaller ones, artisans were at work, and the

things they made were offered for sale in elegant shop-fronts

forming the street-face of the house. Different towns to some

extent specialized in different industries; Silchester had its

dyers, Wroxeter its iron-workers; but in all of them there were

blacksmiths, bronze-workers, joiners, stone-masons, and all the

trades required by the daily life of the town.

These smaller houses were more for use than for luxury; but the larger were not only roomy, but handsome. Their living-

rooms were large and adorned with tessellated pavements and

frescoed walls, and some of them were heated with hypocausts.

Even if they had no upper stories, about which there is no con-

clusive evidence one way or the other, the best had from a dozen

to twenty rooms, of which a good many would be about twenty

feet square. Here is ample space for a rich man's family, with

reception-rooms and bedrooms all on a good scale, and a large

retinue of slaves. In plan these large houses are curiously shape-

less, and their appearance can hardly have been impressive to

the eye of an architect; but they were comfortable, convenient,

spacious; and, surrounded as they were by gardens and opening

on the street through private gates, they were pleasant places

for an easy life.

One general characteristic of these town houses, large and

small alike, is the absence of baths. For the daily bath which

the Roman took before dinner, their inhabitants went to the

public bath-house. Here they met their friends, took their exer-

cise, and went through all the sociable and leisurely ritual of the

Roman bath: the undressing, the cold room, the warm room,

the hot room, and back; the massage and the cold douche, and

so to the dressing-room and home to dine. Sometimes, as in the

later bath-house at Wroxeter, there were two suites of rooms,

doubtless for men and women; as a rule there was only one, and

sometimes an emperor would discover laxities, and issue an edict

that men and women were to bathe at different times. This

happened often enough to show that, in a general way, public

opinion saw no harm in mixed bathing.

Even visitors were provided for. At Silchester there was a

large building which has been identified as an hotel. As strangers

would not be entitled to use the city bath-house, it had a private

one of its own. Doubtless other towns had the same.

Most towns, if not all, had a public water-supply brought by

an aqueduct, not a stone or brick one, striding across country

like those of the drier south, but a covered channel cut in the

ground, tapping a neighbouring stream. When enough had

been provided for the public services, the surplus was available

for private use, though for the most part private houses relied

on their own wells.

For amusements, there was the amphitheatre. Gladiatorial

shows were perhaps too expensive to be common, but beast-

baiting was cheap, and the fondness of Britons for cock-fighting

was already notorious. Verulam was not unique in possessing a

theatre; Tacitus mentions another at Colchester; but Verulam’s

is the only one revealed as yet by the spade. The Verulam

theatre was built about 150. Both from its plan and from what we

know of the drama in Roman times, we can be sure that it was

not used for the production of classical plays; but singers and

dancers may have performed there, and, as no amphitheatre has

been found, it probably served as a bull-ring and cock-pit.

The religious needs of the people were supplied by temples,

of which every town contained a certain number, where gods

of mixed origin were worshipped: partly Celtic, partly Roman,

partly from the east. The temples were small buildings, not

designed for congregations; the ritual was kept going by priests,

and individual worshippers could attend as they pleased, bring-

ing their petitions and their offerings. In every town, too, there

were collegia of the ordinary Roman type: guilds or clubs partly

religious, partly social, partly based on business interests; sub-

ject to legal control and authorization, and thus part of the

public life of the community. One of their functions was to

provide the burial expenses of their members in the cemeteries

which lay along the roads outside the city gates.

Finally, as a body capable of defending itself, the town had its

walls, and could man them at need by calling up its own militia.

Throughout the Antonine period this urban life continued to

flourish. When the forum at Wroxeter was burnt down about

A.D. 160, it was rebuilt. At Verulam private houses were tend-

ing to increase in size and numbers down to the beginning of

the third century. But at some time before the middle of that

century a change began to make itself felt. Evidence has long

been known pointing to a great decline in the prosperity and

population of the towns before the end of the Roman period.

At Silchester, the excavators found that in the best houses people

202

THE TOWNS

had, at some late date, lived squalidly, lighting their fires and doing their cooking upon tessellated pavements. At the time, this date was assumed to fall in the dark years after the departure of the Romans; but now there is no need for such an hypothesis. At Caerwent, a rudely constructed amphitheatre has been built over the deserted ruins of houses inside the town, clear proof that, at a time within the Roman period, the population was much diminished and land within the walls no longer in demand for building.

Further light came with the discovery that at Wroxeter the forum had been again burnt down about A.D. 300 and never rebuilt. The life of the town had not come to an end; the over-thrown columns of the portico, lying half-buried in the street, had been worn on their upper surface by the traffic of long years; but the unrepaired destruction of the forum showed that the trade of the town was ruined, its finances exhausted, and its public life paralysed. For at least a hundred years at the end of its history, Wroxeter was a town economically and politically decayed.

But it was only the Verulam excavations that drew all this together into a coherent and detailed picture. There it was found that by about A.D. 275 the walls were partly ruinous, the theatre was long disused and demolished by service as a common quarry for the town, and the houses were everywhere falling into decay. Verulam ‘must at this time have borne some resemblance to a bombarded city’.1 The vigour of its life was at so low an ebb that no reconstruction was being done. All this indicates a decline of the most far-reaching kind, which had probably made some progress by the middle of the century. The catastrophic manner in which this decline smote a great and central town like Verulam, far away from the frontiers, shows that it was not due to barbarian invasion: its causes, whatever they may have been, were not peripheral, but central.

What, then, were these causes? That is a question which carries us far beyond the confines of British history, and here its obscurities and intricacies can only be dealt with in the briefest outline. First of all, it must be recognized that if we divide the life of Roman Britain into the three heads of urban, rural, and military, the disaster that overwhelmed Verulam affected the first of these alone. Rural life, as we shall see in

1 R. E. M. and T. V. Wheeler, Report on Excavations at Verulam (1936), p. 28.

THEIR DECLINE IN THE THIRD CENTURY

203

the next chapter, continued to prosper: indeed, it prospered increasingly. The defence of the frontier was not impaired.

What happened was that the cities, which until now had stood for romanization and had enjoyed a privilege of civilized life, suffered a blow of the most devastating kind: a blow which left them unable to recover from it.

This did not happen only in Britain. A decay of city life is to be traced in the middle of the third century all over the empire. So general is it, that as lately as 1926, before Verulam had told its story to the excavators, the historian of the economic life of the empire, commenting on the ‘rapid and disastrous decay’ of cities at that time, observed: ‘The only exception seems to have been Britain, where the third century appears to have been a time of peace and prosperity.’1 Something was happening all over the Roman world which was cutting away the foundations of urban civilization: so widespread that it can only have been due to some universal cause, so cumulative that no recovery from it was possible.

Professor Rostovtzeff, who has painted a picture of this decay, using the entire map of the Roman empire as his canvas, traces its origins back to the reign of Septimius Severus. Where earlier emperors had fostered town life as the principal object of their care, Severus openly recognized the army as the basis of his power, and set on foot a movement by which the centre of gravity of the empire’s life was transferred from the town to the camp. The urban aristocracies, instead of being helped and protected, were plundered by a system of taxation and levies that bore upon them with especial severity. The voluntary contributions by which wealthy decurions had always from time to time borne a share in the expenses of the city’s public services were converted into forced levies on capital, and these were developed to such an extent that the position of decurion, instead of being sought as an honour, was one to be avoided as a grievous burden. Everywhere men of substance tried to escape it, and the government found it necessary to make such escape illegal. Caracalla, when in 212 he issued his famous constitutio extending Roman citizenship to a far wider circle, was in effect enlarging the number of persons liable to these burdens; for, while the exact purport of the measure is doubtful, it may for,

1 Rostovtzeff, Social and Economic History of the Roman Empire, p. 422. His chapters ix-xi are fundamental for any discussion of the problems involved.

204

THE TOWNS

have affected the cities only, and in them only the wealthier

classes.1 As the century drew on, the evils of a system by which

emperors raised money and resources in kind through arbitrary

requisition and compulsory levies became more and more in-

tense. While certain classes, notably imperial officials and ten-

ants of imperial estates, were exempt from municipal burdens,

these burdens gradually dissipated the wealth and undermined

the spirit of the propertied classes, and in particular those of

the towns. By the middle of the century the currency had col-

lapsed owing to reckless inflation, and one ephemeral emperor

after another had made matters worse by equipping himself as

best he could through pitiless use of the weapon of compulsory

requisition. Depopulation set in; people left the cities and took

refuge in the woods; and a semblance of prosperity remained

only where the individual household, living on the land, could

produce all it needed for its own subsistence.

That a process of this kind was going on in the third century,

there is no doubt. Its cause, however, is very obscure. Professor

Rostovtzeff has explained it as the effect of a social revolution.

The third-century army, he points out, was in the main an army

of peasants. The Antonine emperors, by raising the status of the

peasantry, had set them on the road towards the development

of class-consciousness; and by the third century, he thinks, the

army had become aware of itself as representing that class, and

had begun to conceive a fixed hatred and envy of the town-

dwellers as a class of persons privileged at their expense, en-

trenched in the enjoyment of a civilized life denied to themselves.

These motives, Rostovtzeff contends, used the emperors as their

tools, and turned the central organization of the empire to the

destruction of that very urban life whose development had

hitherto been its chief function.

Three points may be distinguished in this interpretation of

third-century history. First, that the towns were victimized by

a predatory financial policy which found in their accumulated

wealth an easily tapped source of revenue. There is no doubt

that this was the case, in Britain as elsewhere. Proof of it is to

be found in the tribal inscriptions of Hadrian's Wall, referring,

as we have already seen, to its Severan reconstruction rather

than to its original erection, and indicating that the local

authorities, the urbanized aristocracies of the tribes, had been

1 Rostovtzeff, Social and Economic History of the Roman Empire, p. 369.

called upon to rebuild parts of the Wall at their own expense. Secondly, that this policy indicates a new attitude towards urban civilization, an attitude, if not hostile, at least not decisively favourable. It was no longer a central point in the provincial policy of Rome that city life should be fostered and encouraged. For whatever reason, the towns were no longer an end in themselves. This, too, was evidently the case in Britain. Town life is allowed to decay when the military establishment was being most carefully maintained and when country life was on the up-grade. Obviously, this implies a new scale of values in which less importance is attached to the towns. Thirdly, this change is explained as due to a definite sense of hostility towards the towns on the part of the peasants. On this head the British evidence does not allow us to return a definite verdict. We have no reason to think that either the forces which, about the time of Severus, checked the development of town life in this country, or the forces which by degrees converted this check into a disaster, originated in hostility between the British peasants and the British aristocracy. The actual state of the province, so far as we can detect it, seems to have been peaceful, and the wrecking of town life must have been due to administrative action of a kind universal at that time throughout the empire. This accounts for its violence at Verulam, and its relative slightness in the more outlying parts of the country. In one way Britain suffered less than most parts of the empire: with the exception of Spain it was the only region which never, during the period of anarchy, set up a single pretender to the imperial throne. The Gallic emperors, Postumus, Victorinus, and Tetricus, were unable to arrest the process of decay, and during its temporary independence under Carausius Britain fared no better; but perhaps the Gallic empire and the Carausian period withdrew Britain to some extent from the agonies of the rest, and left it battered and bankrupted, but quiet.

As to its bankruptcy there is no doubt. One of the most remarkable features of the late third century is the numerous and vast hoards of coin that were buried not because the country was unsafe, but because Britain was feeling the effects of the same inflation which multiplied the price of grain in Egypt by 6,000 between the early third century and the time of Diocletian. At Verulam an attempt was made to establish a local currency: a hoard found in the ruins of the theatre contains

206

THE TOWNS

thousands of tiny coins, certainly of local minting, which must for a time have passed current while the inflation was at its worst.

It was in the last years of the century that the revived strength of the central government, due to the work of Diocletian, brought Britain once more under the direct control of Rome.

The visit of Constantius Chlorus in 297 led to a great reorganization of many things in Britain; most of them will be dealt with hereafter, but here one aspect of Constantius's work concerns us.

At Verulam, the end of the third century is marked by a determined attempt to revive the life of the town.

The defences were repaired.

The theatre was rebuilt, and even enlarged.

Old dwelling-houses were restored, and new ones built, some on a large scale.

Whether Verulam represents the rule or the exception we do not know.

At any rate, no similar restoration took place at Wroxeter.

In the case of Verulam, Constantius was reversing the policy of the last hundred years, and instead of bleeding town life was transfusing blood into it.

Probably the new policy was only half-hearted, and applied only to a few of the most important towns.

In any case it failed.

The decay of civic spirit and of faith in the value of urban life had gone too far to be revived by administrative means.

By the middle of the fourth century the effects had worn off.

The greater part of Verulam was uninhabited, a waste of empty land and ruined houses.

Here and there squatters lived among the ruins.

The theatre had become a rubbish-tip, and its orchestra and auditorium were silted up beneath foot upon foot of domestic refuse.

Close round it, a shrunken and impoverished population lived in slum conditions.

There is no reason to think that in this degraded condition fourth-century Verulam was worse off than other towns.

The evidence of squatters' occupation at Silchester and of the deserted and ruined forum at Wroxeter, on the contrary, suggests that it was typical.

But the significance of these facts must not be misunderstood.

They do not imply a break-down or abandonment of the system of local administration.

Shrunken and impoverished though they were, these towns were still the capitals of self-governing tribes.

They still had their permanent decurions and their annual magistrates.

Politically, they were in theory at least, and to some extent in practice, unchanged.

The

FURTHER DECAY IN THE FOURTH CENTURY

207

change was a cultural one. The towns of the first and second century had been the privileged vehicles of Roman civilization. The third century saw that privilege destroyed. Whether it was destroyed by the envy of the unprivileged peasantry, as Rostovtzeff thinks; or because of the helpless rapacity of ephemeral rulers who had no way of supporting themselves except by plundering those whom it was easiest to plunder, the simple and superficial explanation of the historical materialist; or because the deep-rooted Celtic tradition of life, which had in it no place for the town, was not to be overborne by the most earnest efforts of Roman governors; this, at any rate, is certain, that the Britons of the fourth century had given up their earlier belief in, and desire for, the kind of civilization which they had inherited from the Graeco-Roman city-state. The first attempt to romanize British life, by imposing upon it the civilization of the town, had failed. It was when its failure was most complete that the second attempt, based on the spontaneous development of the country house, began to display its most brilliant successes.

XIII

THE COUNTRY-SIDE

Even when the towns were at the height of their prosperity, by far the largest part of the inhabitants of Roman Britain were country-folk, living either in villages or in the isolated farm-houses, large or small, which are called villas ; and of these again, though some worked at industries like mining, pottery, and so forth, the great majority were occupied in agriculture. We should probably not be far wrong if we reckoned that at no time during the Roman period did agriculture occupy less than two-thirds of the inhabitants of Britain.

Something has already been said about agriculture in pre-Roman Britain, and of the development it received at the hands of the Belgae, leading to production on a scale which made it possible to export wheat and other produce. This surplus of production over and above what was needed for the subsistence of the agricultural population was diverted by the Roman conquest into two channels: the imperial taxes, especially the annona or wheat-tax for the maintenance of the army and the government officials, and the local taxes which went to support urban life. The Roman Peace, coupled with these two demands for surplus produce, must have stimulated agriculture from the first, although neither archaeological evidence nor literary sources afford any hint that the Romans taught the Britons to use any new agricultural methods. The agricultural life of the British country-side is in substance and character British in almost every detail known to us; what is Roman about it is the political framework into which it is fitted, and to a certain extent the manners and customs of the people who pursued it.

In the four coloniae, people lived according to Roman custom in towns who were engaged in cultivating their plots of land in the neighbouring country-side. This combination of agriculture with town life may possibly have become general. Many towns were surrounded by regions in which villas and villages are rare. Round Canterbury, for instance, the country seems to have been almost deserted by comparison with the dense villa-population of north-western Kent. This and similar cases elsewhere may indicate that people lived in the towns and worked on the land

Villages

Villas

Roads

400 ft. contours

0

10

20

30

Miles

MAP II. VILLAS AND VILLAGES IN HAMPSHIRE AND WILTSHIRE.

VILLAS AND VILLAGES

209

outside them; but the evidence is not conclusive, and is not borne out by finding agricultural implements in town-houses, except where (as in one case at Silchester) the house was that of a blacksmith or ironmonger. Certainly, too, agriculture was practised by the people dwelling in the civil settlements at military forts, and increasingly so as time went on and the garrisons of the frontier army became more and more like bodies of peasantry. For the most part, however, the agricultural population lived scattered on the land itself, either in villages or in villas.

The distinction between these two types of settlement is of fundamental importance. The villa is a house, large or small, with or without outbuildings, but marked by two characteristics. Economically, it is an isolated farm-house standing in its own land, not clustered with others into a village. Culturally, it is more or less thoroughly romanized in architecture and furniture: it has rectangular rooms, nearly always hypocausts in some of them, generally a suite of baths, and perhaps always painted plaster for decoration on the walls; the things used in it, pottery and so forth, betray a romanized habit of life.

The village is a group of huts, not houses: one-roomed, nearly always circular in shape; sometimes pit-dwellings sunk into the ground, sometimes stone-built structures standing wholly above it; they are never neatly alined on streets or laid out in a regular plan, but clustered shapelessly, generally within some kind of a ditch or fence; and although the people who lived in them used Roman pottery and coins to a certain extent, their daily life was affected by Roman ways very little, and in some cases, especially in the highland zone, not at all.

Seen from the point of view of medieval and modern England, where the distinction between the manor or castle of a land-owner and the cottages of his tenants and labourers is so familiar a fact, this distinction between villa and village naturally suggests a similar relation. But the analogy is misleading. If we look at the relation of villas to villages on the map, instead of finding one villa normally associated with one village or, in the case of a large villa, a group of villages, what we actually find is that cases of such association are very rare, if, indeed, they genuinely exist at all; and that certain regions, so far as we can tell, were exclusively occupied by villas, others exclusively by villages. Thus, Salisbury Plain and Cranborne Chase are

210

THE COUNTRY-SIDE

especially rich in villages, and have been so thoroughly explored

that we may be certain few, if any, villas could have escaped

discovery. Yet, though villas are common on their outskirts,

hardly a single one is known in either district. Other regions,

again, of which the Cotswolds are the most conspicuous instance,

are densely covered with villas but contain no villages. There

are, no doubt, districts containing both; but the possibility of

either existing without the other shows that the relation between

them is something quite different from the relation between

two social grades within one and the same economic system.

We have, in fact, to do with two economic systems existing

side by side. The villa is self-contained. It is the residence of

a Roman or romanized person whom, if we like, we may de-

scribe as a landlord or gentleman; but it does not segregate him

from his social inferiors. Whatever the scale of a villa in size and

in romanization, it always contains enough room for the entire

staff of an estate; and in many cases we can detect certain parts

of its buildings which must have provided quarters for the

labouring population. The village, too, is self-contained in the

sense that whole tracts of agricultural country could be in-

habited by villagers and no one else.

There seems to have been a further distinction in agricultural

methods. Romano-British villages are constantly found to co-

exist with arable land divided up into small fields, irregular in

shape, but as a rule tending towards a square or rectangle. A

vast amount of information on this subject has been collected in

recent years, partly by field-work on the ground, but far more

by air-photography, to which the discovery of these so-called

'Celtic fields' is really due. The result of this very extensive

work has been to establish two generalizations: that Romano-

British villages, broadly speaking, always have fields of this kind,

and that villas never have them. The difference between village-

economy and villa-economy thus extended to the lay-out of their

arable land. The villas cultivated large open fields, the villages

small enclosed ones.

This suggests a difference in tenure. The undivided arable

land of the villa points to undivided ownership or tenancy; the

small fields of a village suggest that its arable land was divided

into parcels held, on whatever tenure, by individual villagers.

The two systems might conceivably rest on an original distinc-

tion between land owned by individuals and land owned in

TWO AGRICULTURAL SYSTEMS

211

common by villages and parcelled out from time to time among

their members.

It also suggests a difference in agricultural technique. All

over the Roman world, people used various types of the

light wheelless plough which the Romans called aratrum; but

here and there they found the Celtic peoples using the heavy

wheeled plough, invented, as Pliny says, in Raetia, which the

Gauls called caruca. The essence of the invention was that by

supporting it on wheels the cultivator could use a much heavier

implement with a much higher penetrating power. Even on

light soils this gave the caruca an advantage, for it could do at

one operation as much work as the aratrum, by cross-ploughing,

could do in two; but also the caruca could be used on heavier

soils where the aratrum could not be used at all. To utilize its

penetrating power to the full, the wheeled plough was fitted

with a coulter or vertical knife in front of the share, to cut the

soil in preparation for turning a sod, and a share designed to

lift the sod sideways so that the mould-board could turn it clean

over, whereas the aratrum in its simplest form did little but grub

the soil like a pig’s snout and heap it into a ridge on either side

of the furrow.

Of these two types of plough, the aratrum would be the more

suitable for the small fields of village-arable, which are seldom

more than an acre or two in extent, and whose peasant culti-

vators must have used plant of a cheap and simple kind. The

open fields of a villa, on the contrary, would suit the more

powerful and expensive caruca, especially in view of the fact that

villas are sometimes situated on heavy soils where the aratrum

would be useless. In fact, it has lately been pointed out that in

France both kinds of plough have survived into modern times,

together with two types of field-system: the araire being co-

extensive with systems of plots resembling the ‘Celtic fields’ of

Britain, the charrue with open fields divided into strips.1

Both aratrum and caruca were used in Roman Britain. The

aratrum is represented in the little bronze statuette of a cloaked

and hooded British ploughman found at Piercebridge, in the

county of Durham, and now in the British Museum; it shows an

implement (no doubt the ordinary Romano-British plough)

whose action was of the grubbing kind described above.

Large coulters clearly belonging to a caruca have been found not

1 Marc Bloch, Les Caractères originaux de l’histoire rurale française (Oslo, 1931).

212

THE COUNTRY-SIDE

only at pre-Roman Belgic sites in Britain, but also at Roman

sites: Silchester, Great Chesterford in Essex, Witcombe in Glou-

cestershire; so that the heavy coulter-plough, which was intro-

duced into the country, as it seems, by the Belgae, certainly

played its part in Romano-British agriculture.1

Can we go a step farther, and argue that the aratrum was used

by the villagers, the caruca by the villa-dwellers? The evidence

is in favour of such an hypothesis but is not sufficient to prove

it. The Piercebridge model was found far away from the

nearest villa: the building at Great Chesterford where coulters

were found is a typical villa, and so is the house at Witcombe.

The Silchester coulters, being found not on the land but in

a town, tell us nothing definite. The Belgae, according to

Caesar, lived in isolated farms; but they did not introduce the

isolated farm into Britain; it existed there before they came.

Perhaps what they did was, by introducing the caruca, which

was adapted for use in open fields and not in little enclosed plots,

to give a new impetus to open-field cultivation, which for them

was associated with the existence of isolated farms, and thus

enable the agriculture of the isolated farm to outdistance its

rival, the agriculture of the villages, tied as that was to the

system of cultivation by small plots. And at the same time the

association of the caruca with the isolated farm made it possible

for isolated farms to encroach on forest-land and begin culti-

vating heavy soils with which the aratrum could not cope.

At the Claudian conquest, therefore, it seems likely (one can

say no more; the evidence being so slender as it is) that two

systems of land-tenure and of cultivation existed side by side,

corresponding with two types of settlement. There was the

village, with its communal land divided up into little fields

parcelled out among the members of the commune and culti-

vated with the light plough, and there was the isolated family

farm, with its open fields, cultivated, where Belgic influence had

made itself felt, with the coulter-plough, and tending here and

there to encroach on the heavier soils of the forest-land. It was

thus already becoming clear that, of the two systems, one had

in it germs of progress, the other was relatively stagnant. The

village-economy was rigid and unprogressive because the minute

subdivision of land and the smallness of the capital commanded

by any individual made reforms all but impossible. The more

1 Karslake in Antiquaries Journal, xiii. 455; Hawkes in Antiquity, ix (1935), p. 339.

ORIGIN OF THE VILLA-SYSTEM

213

substantial and independent farmers, the class qualified by

wealth and circumstances to embrace new movements, were the

dwellers in the isolated farms. These independent farmers, after

the conquest, by degrees acquired Roman ways of living, and

their farms became Roman villas; the villages were too poor

and too unprogressive to undergo any such change, and they

retained their primitive character.

If it is true, as the above account implies, that what we may

call the villa-system in Roman Britain was the romanized ver-

sion of a system of land-tenure and cultivation already existing

before the conquest, and in especial stimulated and developed

by the Belgic settlement, there ought to be some positive evi-

dence that such a system existed before the Romans came.

There is such evidence; but it is scanty, and still needs to be

confirmed and supplemented by scientific excavation at villa-

sites in the south-eastern counties of England.

In the first place, what we know of Celtic economic life sug-

gests that isolated farms must have been quite common in

pre-Roman Britain. And this is borne out both by Caesar’s

statement that he saw them in Kent,1 and by the fact that

isolated farms exist to this day in Ireland, Wales, Cornwall, and

Scotland, the parts of the British Isles that were never either

romanized or saxonized, at least as commonly as elsewhere,

and traceable back to a remote antiquity. It does not follow,

of course, that they were all associated with open-field culti-

vation.

Further, there is archaeological evidence for the existence of

pre-Roman isolated farms in the lowland zone; one dating from

the Bronze Age has recently been dug on Harrow Hill near

Worthing in Sussex: evidence, too, that such farms existed in

the Roman period in an un-romanized shape; an example has

been dug on Rockbourne Down in Wiltshire. In the highland

zone, un-romanized isolated farms of the Roman and pre-

Roman periods are common. And lastly, there are several

villas in the south-eastern counties where excavation has re-

vealed traces of pre-Roman occupation on the site, either cer-

tainly or presumably belonging to a pre-Roman farmstead:

examples are at Otford (Kent), Newport (Isle of Wight), Knowl

Hill (Berkshire), and perhaps Ashtead (Surrey) and Folkestone

(Kent). The classical example is at Mayen in the Eifel, where

1 See above, p. 28.

2720.1

P

214

THE COUNTRY-SIDE

a complete farm of the La Tène period has been studied in

all the phases of its conversion into an ordinary Roman villa.

Nothing so conclusive has been found as yet in Britain, but the

British evidence, slight though it is, suffices to justify the assump-

tion that the principle proved for Germany holds good for

Britain also.

It does not, of course, follow that all Roman villas occupy the

sites of pre-Roman farms. On the contrary, if the farm-economy

was more efficient and progressive than village-economy, and

in particular better adapted to the tillage of cleared forest-land,

extensions of cultivation during the Roman period would largely

come about through the planting of new villas on virgin sites, in

forest-clearings and elsewhere. There is evidence that some-

thing of this kind actually happened. In the Cambridge district

a few villas were placed in forest-clearings where no earlier finds

occur; the same thing happens on the edge of the forest in

southern Berkshire. That this development took place, or at

any rate began, at an early date is proved by the villa lately

dug at Ditchley, near Oxford. It stood in a clearing of Wych-

wood forest, and excavation has shown that in its first form, as

a timber house of Roman type, it dates from at least the Flavian

age, if not from the reign of Nero. The Roman period also saw

some extension of village-economy, but this will be considered

later.

The history of the villa-system in Britain is still very obscure,

owing to the fewness of the sites on which really scientific

digging has been done. We know a great deal about the archi-

tecture and furnishings of these villas at the time of their greatest

prosperity, but it is very seldom either that this time can be

accurately dated or that search has been made, underneath its

remains, for those of earlier buildings on the same site. From

the evidence of coins picked up on villa-sites it is abundantly

certain that most of them were flourishing in the fourth century;

excavation has proved that some were first built at this time,

others in the third century, others in the second. Broadly speak-

ing, the villa-system appears to have been growing, slowly but

without a check, from the first century to the fourth. In Kent, on

the Hampshire coast, and in the Isle of Wight there was already

in the late first century a considerable development of villa life,

which spread, before the century was over, as far north as the

Thames valley and as far west as Somerset. Evidently the richer

DEVELOPMENT OF THE VILLA-SYSTEM

215

land-owners began rebuilding their country houses after a romanized pattern quite soon after the conquest.

The building of villas may be regarded as a second movement of romanization, distinct from that of the towns, and having certain special features of its own. Its most remarkable feature is its spontaneity. It was not due to an influx of settlers from abroad taking up land in Britain. Partly that is shown by the continuity between villas and pre-Roman buildings existing on the same sites, which attests the reconstruction of old farmsteads in a new style; partly by the entirely British type of skulls found at villa-sites ; partly by general historical considerations, such as the improbability of foreign settlers at so early a date going anywhere except to the towns. For these and similar reasons no student of the subject is disposed to reject the statement of Haverfield that ‘the inhabitants of the villas were, so far as we can tell, romanized natives of Britain. Here and there a settler from Gaul, or even from Italy, may have found his way to the island and acquired landed estate. There is, however, a singular absence of proof or indication of any such thing. We seem clearly to be dealing with an indigenous native population.’1

This movement may have been encouraged by government policy, offering easy terms to settlers on uninhabited land; but in great part it must have been spontaneous. It was no doubt connected with the movement of urban romanization, and in one sense a reflection of it; for the people who first converted their farms into villas were very likely just those wealthy Britons who had a seat in the ordo of their tribe and a house in the tribal capital, and learnt there to live in the Roman style; but it indicates a genuine spontaneity, because it shows how these persons wished to live, when they were no longer performing their official functions under the eye of the governor and under the pressure of his known policy.

At the same time, it was confined to one class. It was a privilege of the well-to-do, and the peasants of the villages did not share in it. In this way also it differed from the romanization of the towns, which affected rich and poor alike.

Another difference is revealed by its history. Partly because of its spontaneity, partly because of its diffusion, it possessed a staying power which the urban romanization lacked. Nothing can be more striking than the contrast between the collapse of

1 Roman Occupation of Britain, p. 232.

216

THE COUNTRY-SIDE

the towns in the third century, with the complete failure of

every effort to redeem it, and the continued and increasing

prosperity of the villas. When Haverfield described the late

third and fourth centuries as the golden age of Roman Britain,

he had in mind the evidences of this widespread rural prosperity

and civilization. The evidence of decay in the towns had not

yet come to hand; when it did, the prosperity of the villas only

appeared more striking by contrast.

Haverfield went so far as to suggest that it was the reign

of Septimius Severus which, in particular, marked a forward

movement in the civilization of the villas. The evidence on

which that suggestion is based, though never amounting to

proof, is still valid to-day. We have seen that it was precisely

the Severan age that saw the first check to the rising tide of

urban prosperity. The two things can hardly be unrelated. If

it is true that the early third century was the time when system-

atic pillage of the cities ‘dealt a mortal blow at the civic spirit

of the higher classes, and induced them to conceal their wealth

and appear as poor as possible’,1 these same causes would lead

the wealthier land-owners to spend more time and money on

their estates, and to improve their country houses rather than

keep up their town houses.

Once this movement of land-owners back to their estates had

begun, other causes contributed to accentuate it. Before long,

the currency began to depreciate. The consequent rise and

uncertainty of prices made town life unpleasantly precarious.

Not only was trade ruined, but people living in towns must

have had difficulty in procuring food and necessaries of every

kind. In the country, where most things that a household re-

quired were produced on the estate, people could live on their

own granaries and store-rooms, unaffected by fluctuations in

the value of money. Before the end of the third century the

wealthier classes had learned to think of their villas as places

where they could live in a degree of comfort and security not

possible in the towns. The habit of country life, once reaffirmed

and strengthened by such inducements, was not to be broken.

The attempt of Constantius to re-establish the town as the centre

of civilized life beat against it in vain. The Roman civilization

of Britain had left the towns and transplanted itself to the

villas.

1 Rostovtzeff, Social and Economic History of the Roman Empire, p. 400.

DISTRIBUTION OF VILLAS

217

The extent and distribution of this rural civilization can be studied on the map with a considerable degree of accuracy, for Roman villas have so long been an object of interest to people of all kinds in England that records of their discovery are voluminous and, on the whole, trustworthy. We know about 500 of them. Of these 90 per cent. lie not only in the lowland zone but south-east of a line joining the Severn and Trent. Even within this restricted area there are certain districts where they hardly occur. They are densest in north-west Kent round Maidstone; west Sussex; the Hampshire basin east of Salisbury and south of Silchester; the Somerset-Dorset borders round Ilchester; the Avon basin round Bath; and the Cotswold plateau round Cirencester, with a belt of them extending thence along the oolite-ridge through Northamptonshire into Lincolnshire. Their absence from the Weald, the New Forest, and the clay belt joining Oxford to the Fens, is due to a total scarcity of population in those districts. On the other hand, their rarity in Cranborne Chase, Salisbury Plain, and the middle Thames valley is connected with the fact that here the village-population is extremely dense. Their relative scarcity in East Anglia seems to be connected with the first cause rather than the second.

Outside this area there are one or two in Cornwall, a fair number in the extreme east and south of Wales, a very few in the midlands, especially near Leicester, and over a dozen in Yorkshire.

It has already been remarked that in some cases they have been placed on heavier soils than were commonly tilled before the Roman period. But these cases are very exceptional. Most villas were built upon the lighter soils, chalk and oolite and gravel, which needed little clearing and no drainage. As compared with villages, however, they tend to select a rather special type of site: a valley-slope facing south or east, not too high up, with shelter from wind, exposure to the sun, and water close at hand. The villages of Salisbury Plain lie out on the wind-swept summit of the plateau; the villas of the Cotswolds nestle warmly on the sides of the valleys, basking in the sunshine.

In ground-plan, the great majority belong to one pattern: a row of rooms all entered from a corridor or veranda running along the front, with a wing jutting forward at each end. Often there is a corridor along the back as well; sometimes it runs all round. The front door is in the centre of the corridor, and leads

218

THE COUNTRY-SIDE

through it into a central room which might be described in

house-agents' language as the 'lounge hall'. In a house of

average size there will be two or three other rooms on each side

of this, and the best living-rooms will be in the wings. Such a

plan would give seven or nine rooms as a normal number; but

some villas of the same pattern are smaller, many are larger,

and some have begun on this scale but have had additions made

to them. Whether they ever had upper stories is not known;

probably as a rule they did not. They were generally built in

timber and plaster on a stone foundation, with tiled or slated

roofs.

Houses of this kind are as common in Gaul and Germany as

they are in Britain, and scholars have differed as to whether the

type is a Celtic invention or an importation from the Mediter-

ranean world. Neither alternative seems quite to fit the facts.

The partisans of a Celtic origin cannot point to such houses

existing in Celtic lands before the Roman conquest; those who

think that it came from the Mediterranean cannot produce

satisfying parallels from the architecture of the south. It seems

most probable that the type was invented in the Celtic regions

under Roman rule and influence, and should be regarded as a

testimony (not the only one) to the existence of a specifically

Romano-Celtic civilization.

A villa of this kind has certain accessories which, except per-

haps in the very poorest, were regarded as indispensable. A

rectangular, walled farm-yard lay in front of it, and generally

surrounded it on all sides. This farm-yard contained other

buildings: barns, stables, and quarters for the farm labourers.

The house proper was thus reserved for the master and his

family, no doubt including domestic slaves. There was always,

except in the poorest villas, a suite of baths, either in the house

itself or forming a separate building. A well in the yard supplied

water, unless, as was sometimes the case, it was laid on by means

of an aqueduct; and sometimes it is possible to recognize a

threshing-floor, a dovecote, or the circular track of a horse-

mill.

Among the humbler villas, a certain number belong to a quite

different type of house. This is the so-called basilican house, or,

more simply, the barn-dwelling. It is an oblong building with

two rows of posts running along its length and dividing it into

a nave and two aisles, exactly like the barns which are still to be